Training in progress, step 1250, checkpoint
Browse files- checkpoint-1250/README.md +202 -0
- checkpoint-1250/adapter_config.json +34 -0
- checkpoint-1250/adapter_model.safetensors +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt +3 -0
- checkpoint-1250/global_step1250/mp_rank_00_model_states.pt +3 -0
- checkpoint-1250/latest +1 -0
- checkpoint-1250/rng_state_0.pth +3 -0
- checkpoint-1250/rng_state_1.pth +3 -0
- checkpoint-1250/rng_state_2.pth +3 -0
- checkpoint-1250/rng_state_3.pth +3 -0
- checkpoint-1250/rng_state_4.pth +3 -0
- checkpoint-1250/rng_state_5.pth +3 -0
- checkpoint-1250/rng_state_6.pth +3 -0
- checkpoint-1250/rng_state_7.pth +3 -0
- checkpoint-1250/scheduler.pt +3 -0
- checkpoint-1250/special_tokens_map.json +30 -0
- checkpoint-1250/tokenizer.json +0 -0
- checkpoint-1250/tokenizer_config.json +133 -0
- checkpoint-1250/trainer_state.json +2308 -0
- checkpoint-1250/training_args.bin +3 -0
- checkpoint-1250/zero_to_fp32.py +674 -0
checkpoint-1250/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: microsoft/Phi-3-mini-4k-instruct
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.14.0
|
checkpoint-1250/adapter_config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"eva_config": null,
|
| 7 |
+
"exclude_modules": null,
|
| 8 |
+
"fan_in_fan_out": false,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": true,
|
| 11 |
+
"layer_replication": null,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 16,
|
| 16 |
+
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.0,
|
| 18 |
+
"megatron_config": null,
|
| 19 |
+
"megatron_core": "megatron.core",
|
| 20 |
+
"modules_to_save": null,
|
| 21 |
+
"peft_type": "LORA",
|
| 22 |
+
"r": 8,
|
| 23 |
+
"rank_pattern": {},
|
| 24 |
+
"revision": null,
|
| 25 |
+
"target_modules": [
|
| 26 |
+
"down_proj",
|
| 27 |
+
"qkv_proj",
|
| 28 |
+
"gate_up_proj",
|
| 29 |
+
"o_proj"
|
| 30 |
+
],
|
| 31 |
+
"task_type": "CAUSAL_LM",
|
| 32 |
+
"use_dora": false,
|
| 33 |
+
"use_rslora": false
|
| 34 |
+
}
|
checkpoint-1250/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d80f2c648a0aca0e72371b2c69b66f6dc19cfe220178a7db24f7be8e742472e
|
| 3 |
+
size 25200088
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2fd59f44c88435d7c052214d443456188912be166c8fc5f2617d7fc6de7158e
|
| 3 |
+
size 18881328
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:749c67d171d435c64243ccb70774f27c5c54127f62f0bd2e5e836b7851202699
|
| 3 |
+
size 18881328
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f79c3bee082f39b9d4bec49cdbd6ed89ca5a903ef2a21e3c17902232e1995275
|
| 3 |
+
size 18881328
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc1a93306f627fa2928f9bf17d905f430c6e94d4308189f6110e0a34b7a36838
|
| 3 |
+
size 18881392
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d730538d430613b27b32f8a15100aff6033cdf54f71d6e81e62c0cde1a332661
|
| 3 |
+
size 18881392
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d46734a9137c52a4ae65eb8e8de509f99187425fd4e8a1513fef49b8116a21fa
|
| 3 |
+
size 18881392
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:536d3d7dc28599c41c2f7d6e615beaae3b4adf794414f8ab6de4eb1d7261c293
|
| 3 |
+
size 18881392
|
checkpoint-1250/global_step1250/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f48ad621149e0ebd8311d53c4609508e3029a3f05504225e707f64b253e0f71
|
| 3 |
+
size 18881392
|
checkpoint-1250/global_step1250/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b20d0b9de290665533879d72b4b873fd4bcbe3fa9f814e23621bbc8c8169894b
|
| 3 |
+
size 25379244
|
checkpoint-1250/latest
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
global_step1250
|
checkpoint-1250/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:343f4286599993ed6196c5401702754136843f36dd00127e9d480dcd80159fe3
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9befffc75eb87b0bc636a17f92325765a402aea6df9004d264ae182c4b3348b
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71f1a9ba48b7fb3c3e4f1e2cbee76fc776168ccecb572d29900e58fb5dabf633
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e10b5290c005590436e0e89a12ce21b3888b377bb74aaf05b5f52d01666c809d
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18ad35928e22a674bcbecc7be68aedfab901d49755cd3b1ebaba37e489694d1e
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8fc0855abe8ff42d92eb2c414b25f3a9ba018466841df3ced77b7e05b590df0b
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a40782745bd53ffde678e156f981066d4bc6322666acf70c1894052c32a94fd8
|
| 3 |
+
size 15984
|
checkpoint-1250/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8622c198dc2441f20ba21ffcfa5ac91baee63fb99abb6137c373932a6870ba75
|
| 3 |
+
size 15984
|
checkpoint-1250/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85a32e7d79d4bce24e8039d66138f2f6eb81d15d973d067af5d2c4296fc1dd0b
|
| 3 |
+
size 1064
|
checkpoint-1250/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|end|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
checkpoint-1250/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1250/tokenizer_config.json
ADDED
|
@@ -0,0 +1,133 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": true,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": false
|
| 29 |
+
},
|
| 30 |
+
"32000": {
|
| 31 |
+
"content": "<|endoftext|>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
},
|
| 38 |
+
"32001": {
|
| 39 |
+
"content": "<|assistant|>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": true,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": true
|
| 45 |
+
},
|
| 46 |
+
"32002": {
|
| 47 |
+
"content": "<|placeholder1|>",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": false,
|
| 50 |
+
"rstrip": true,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": true
|
| 53 |
+
},
|
| 54 |
+
"32003": {
|
| 55 |
+
"content": "<|placeholder2|>",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": false,
|
| 58 |
+
"rstrip": true,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": true
|
| 61 |
+
},
|
| 62 |
+
"32004": {
|
| 63 |
+
"content": "<|placeholder3|>",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": false,
|
| 66 |
+
"rstrip": true,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": true
|
| 69 |
+
},
|
| 70 |
+
"32005": {
|
| 71 |
+
"content": "<|placeholder4|>",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": false,
|
| 74 |
+
"rstrip": true,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": true
|
| 77 |
+
},
|
| 78 |
+
"32006": {
|
| 79 |
+
"content": "<|system|>",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": false,
|
| 82 |
+
"rstrip": true,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": true
|
| 85 |
+
},
|
| 86 |
+
"32007": {
|
| 87 |
+
"content": "<|end|>",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": false,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": true
|
| 93 |
+
},
|
| 94 |
+
"32008": {
|
| 95 |
+
"content": "<|placeholder5|>",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": false,
|
| 98 |
+
"rstrip": true,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": true
|
| 101 |
+
},
|
| 102 |
+
"32009": {
|
| 103 |
+
"content": "<|placeholder6|>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": true,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": true
|
| 109 |
+
},
|
| 110 |
+
"32010": {
|
| 111 |
+
"content": "<|user|>",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": false,
|
| 114 |
+
"rstrip": true,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": true
|
| 117 |
+
}
|
| 118 |
+
},
|
| 119 |
+
"bos_token": "<s>",
|
| 120 |
+
"chat_template": "{% set system_message = 'You are a helpful AI assistant.' %}{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ '<s>' + '<|system|>\n' + system_message + '<|end|>\n' }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|user|>\n' + content + '<|end|>\n<|assistant|>\n' }}{% elif message['role'] == 'assistant' %}{{ content + '<|end|>' + '\n' }}{% endif %}{% endfor %}",
|
| 121 |
+
"clean_up_tokenization_spaces": false,
|
| 122 |
+
"eos_token": "<|end|>",
|
| 123 |
+
"extra_special_tokens": {},
|
| 124 |
+
"legacy": false,
|
| 125 |
+
"model_max_length": 4096,
|
| 126 |
+
"pad_token": "<|endoftext|>",
|
| 127 |
+
"padding_side": "right",
|
| 128 |
+
"sp_model_kwargs": {},
|
| 129 |
+
"split_special_tokens": false,
|
| 130 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 131 |
+
"unk_token": "<unk>",
|
| 132 |
+
"use_default_system_prompt": false
|
| 133 |
+
}
|
checkpoint-1250/trainer_state.json
ADDED
|
@@ -0,0 +1,2308 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.1101243339253997,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 1250,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008880994671403197,
|
| 13 |
+
"grad_norm": 0.04571289196610451,
|
| 14 |
+
"learning_rate": 4.999451708687114e-06,
|
| 15 |
+
"logits/chosen": 14.56671142578125,
|
| 16 |
+
"logits/rejected": 15.112574577331543,
|
| 17 |
+
"logps/chosen": -0.26506316661834717,
|
| 18 |
+
"logps/rejected": -0.3439488410949707,
|
| 19 |
+
"loss": 0.9267,
|
| 20 |
+
"rewards/accuracies": 0.574999988079071,
|
| 21 |
+
"rewards/chosen": -0.39759472012519836,
|
| 22 |
+
"rewards/margins": 0.11832849681377411,
|
| 23 |
+
"rewards/rejected": -0.5159232020378113,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.017761989342806393,
|
| 28 |
+
"grad_norm": 0.0512714721262455,
|
| 29 |
+
"learning_rate": 4.997807075247147e-06,
|
| 30 |
+
"logits/chosen": 14.376543045043945,
|
| 31 |
+
"logits/rejected": 14.862703323364258,
|
| 32 |
+
"logps/chosen": -0.2708089351654053,
|
| 33 |
+
"logps/rejected": -0.32412824034690857,
|
| 34 |
+
"loss": 0.936,
|
| 35 |
+
"rewards/accuracies": 0.5,
|
| 36 |
+
"rewards/chosen": -0.4062133729457855,
|
| 37 |
+
"rewards/margins": 0.07997899502515793,
|
| 38 |
+
"rewards/rejected": -0.4861923158168793,
|
| 39 |
+
"step": 20
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.02664298401420959,
|
| 43 |
+
"grad_norm": 0.058383647352457047,
|
| 44 |
+
"learning_rate": 4.9950668210706795e-06,
|
| 45 |
+
"logits/chosen": 14.208717346191406,
|
| 46 |
+
"logits/rejected": 15.370651245117188,
|
| 47 |
+
"logps/chosen": -0.28206294775009155,
|
| 48 |
+
"logps/rejected": -0.38387423753738403,
|
| 49 |
+
"loss": 0.9215,
|
| 50 |
+
"rewards/accuracies": 0.625,
|
| 51 |
+
"rewards/chosen": -0.42309442162513733,
|
| 52 |
+
"rewards/margins": 0.15271687507629395,
|
| 53 |
+
"rewards/rejected": -0.5758112668991089,
|
| 54 |
+
"step": 30
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.035523978685612786,
|
| 58 |
+
"grad_norm": 0.06262075155973434,
|
| 59 |
+
"learning_rate": 4.9912321481237616e-06,
|
| 60 |
+
"logits/chosen": 14.768765449523926,
|
| 61 |
+
"logits/rejected": 15.169331550598145,
|
| 62 |
+
"logps/chosen": -0.27857059240341187,
|
| 63 |
+
"logps/rejected": -0.3388269543647766,
|
| 64 |
+
"loss": 0.9386,
|
| 65 |
+
"rewards/accuracies": 0.5,
|
| 66 |
+
"rewards/chosen": -0.4178559184074402,
|
| 67 |
+
"rewards/margins": 0.09038447588682175,
|
| 68 |
+
"rewards/rejected": -0.5082404017448425,
|
| 69 |
+
"step": 40
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.04440497335701599,
|
| 73 |
+
"grad_norm": 0.06259036809206009,
|
| 74 |
+
"learning_rate": 4.986304738420684e-06,
|
| 75 |
+
"logits/chosen": 14.950456619262695,
|
| 76 |
+
"logits/rejected": 15.232122421264648,
|
| 77 |
+
"logps/chosen": -0.2961367070674896,
|
| 78 |
+
"logps/rejected": -0.3322262465953827,
|
| 79 |
+
"loss": 0.9317,
|
| 80 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 81 |
+
"rewards/chosen": -0.44420504570007324,
|
| 82 |
+
"rewards/margins": 0.054134320467710495,
|
| 83 |
+
"rewards/rejected": -0.4983394145965576,
|
| 84 |
+
"step": 50
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.04440497335701599,
|
| 88 |
+
"eval_logits/chosen": 14.56529426574707,
|
| 89 |
+
"eval_logits/rejected": 14.895020484924316,
|
| 90 |
+
"eval_logps/chosen": -0.2806546986103058,
|
| 91 |
+
"eval_logps/rejected": -0.3486972451210022,
|
| 92 |
+
"eval_loss": 0.9381324052810669,
|
| 93 |
+
"eval_rewards/accuracies": 0.5274725556373596,
|
| 94 |
+
"eval_rewards/chosen": -0.4209820330142975,
|
| 95 |
+
"eval_rewards/margins": 0.10206379741430283,
|
| 96 |
+
"eval_rewards/rejected": -0.5230458974838257,
|
| 97 |
+
"eval_runtime": 25.2574,
|
| 98 |
+
"eval_samples_per_second": 28.823,
|
| 99 |
+
"eval_steps_per_second": 3.603,
|
| 100 |
+
"step": 50
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.05328596802841918,
|
| 104 |
+
"grad_norm": 0.07301533967256546,
|
| 105 |
+
"learning_rate": 4.980286753286196e-06,
|
| 106 |
+
"logits/chosen": 14.195574760437012,
|
| 107 |
+
"logits/rejected": 15.173194885253906,
|
| 108 |
+
"logps/chosen": -0.2693648636341095,
|
| 109 |
+
"logps/rejected": -0.33997970819473267,
|
| 110 |
+
"loss": 0.9319,
|
| 111 |
+
"rewards/accuracies": 0.512499988079071,
|
| 112 |
+
"rewards/chosen": -0.40404725074768066,
|
| 113 |
+
"rewards/margins": 0.10592226684093475,
|
| 114 |
+
"rewards/rejected": -0.5099694728851318,
|
| 115 |
+
"step": 60
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.06216696269982238,
|
| 119 |
+
"grad_norm": 0.0659889206290245,
|
| 120 |
+
"learning_rate": 4.973180832407471e-06,
|
| 121 |
+
"logits/chosen": 14.910173416137695,
|
| 122 |
+
"logits/rejected": 15.361429214477539,
|
| 123 |
+
"logps/chosen": -0.28456225991249084,
|
| 124 |
+
"logps/rejected": -0.3702812194824219,
|
| 125 |
+
"loss": 0.9185,
|
| 126 |
+
"rewards/accuracies": 0.625,
|
| 127 |
+
"rewards/chosen": -0.42684346437454224,
|
| 128 |
+
"rewards/margins": 0.12857840955257416,
|
| 129 |
+
"rewards/rejected": -0.5554218292236328,
|
| 130 |
+
"step": 70
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.07104795737122557,
|
| 134 |
+
"grad_norm": 0.05815625935792923,
|
| 135 |
+
"learning_rate": 4.964990092676263e-06,
|
| 136 |
+
"logits/chosen": 14.407182693481445,
|
| 137 |
+
"logits/rejected": 14.948204040527344,
|
| 138 |
+
"logps/chosen": -0.292889267206192,
|
| 139 |
+
"logps/rejected": -0.3381648063659668,
|
| 140 |
+
"loss": 0.9388,
|
| 141 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 142 |
+
"rewards/chosen": -0.43933385610580444,
|
| 143 |
+
"rewards/margins": 0.06791339069604874,
|
| 144 |
+
"rewards/rejected": -0.5072472095489502,
|
| 145 |
+
"step": 80
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.07992895204262877,
|
| 149 |
+
"grad_norm": 0.06627190113067627,
|
| 150 |
+
"learning_rate": 4.9557181268217225e-06,
|
| 151 |
+
"logits/chosen": 14.622471809387207,
|
| 152 |
+
"logits/rejected": 15.167770385742188,
|
| 153 |
+
"logps/chosen": -0.28155821561813354,
|
| 154 |
+
"logps/rejected": -0.33633899688720703,
|
| 155 |
+
"loss": 0.9256,
|
| 156 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 157 |
+
"rewards/chosen": -0.4223373532295227,
|
| 158 |
+
"rewards/margins": 0.08217118680477142,
|
| 159 |
+
"rewards/rejected": -0.5045084953308105,
|
| 160 |
+
"step": 90
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.08880994671403197,
|
| 164 |
+
"grad_norm": 0.0724545568227768,
|
| 165 |
+
"learning_rate": 4.9453690018345144e-06,
|
| 166 |
+
"logits/chosen": 14.289724349975586,
|
| 167 |
+
"logits/rejected": 14.882037162780762,
|
| 168 |
+
"logps/chosen": -0.2791440486907959,
|
| 169 |
+
"logps/rejected": -0.35329627990722656,
|
| 170 |
+
"loss": 0.9374,
|
| 171 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 172 |
+
"rewards/chosen": -0.41871610283851624,
|
| 173 |
+
"rewards/margins": 0.11122839152812958,
|
| 174 |
+
"rewards/rejected": -0.5299445390701294,
|
| 175 |
+
"step": 100
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.08880994671403197,
|
| 179 |
+
"eval_logits/chosen": 14.337930679321289,
|
| 180 |
+
"eval_logits/rejected": 14.689269065856934,
|
| 181 |
+
"eval_logps/chosen": -0.2726942300796509,
|
| 182 |
+
"eval_logps/rejected": -0.34668418765068054,
|
| 183 |
+
"eval_loss": 0.9302808046340942,
|
| 184 |
+
"eval_rewards/accuracies": 0.5384615659713745,
|
| 185 |
+
"eval_rewards/chosen": -0.40904131531715393,
|
| 186 |
+
"eval_rewards/margins": 0.11098497360944748,
|
| 187 |
+
"eval_rewards/rejected": -0.5200263261795044,
|
| 188 |
+
"eval_runtime": 25.2585,
|
| 189 |
+
"eval_samples_per_second": 28.822,
|
| 190 |
+
"eval_steps_per_second": 3.603,
|
| 191 |
+
"step": 100
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.09769094138543517,
|
| 195 |
+
"grad_norm": 0.08156246691942215,
|
| 196 |
+
"learning_rate": 4.933947257182901e-06,
|
| 197 |
+
"logits/chosen": 14.499124526977539,
|
| 198 |
+
"logits/rejected": 14.916313171386719,
|
| 199 |
+
"logps/chosen": -0.2798352837562561,
|
| 200 |
+
"logps/rejected": -0.3477734327316284,
|
| 201 |
+
"loss": 0.9243,
|
| 202 |
+
"rewards/accuracies": 0.5625,
|
| 203 |
+
"rewards/chosen": -0.4197530150413513,
|
| 204 |
+
"rewards/margins": 0.10190720856189728,
|
| 205 |
+
"rewards/rejected": -0.5216602087020874,
|
| 206 |
+
"step": 110
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.10657193605683836,
|
| 210 |
+
"grad_norm": 0.08161844313144684,
|
| 211 |
+
"learning_rate": 4.921457902821578e-06,
|
| 212 |
+
"logits/chosen": 13.595013618469238,
|
| 213 |
+
"logits/rejected": 14.390353202819824,
|
| 214 |
+
"logps/chosen": -0.26682502031326294,
|
| 215 |
+
"logps/rejected": -0.3336995542049408,
|
| 216 |
+
"loss": 0.9123,
|
| 217 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 218 |
+
"rewards/chosen": -0.400237500667572,
|
| 219 |
+
"rewards/margins": 0.10031183809041977,
|
| 220 |
+
"rewards/rejected": -0.5005493760108948,
|
| 221 |
+
"step": 120
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.11545293072824156,
|
| 225 |
+
"grad_norm": 0.28624778985977173,
|
| 226 |
+
"learning_rate": 4.907906416994146e-06,
|
| 227 |
+
"logits/chosen": 13.711044311523438,
|
| 228 |
+
"logits/rejected": 14.558542251586914,
|
| 229 |
+
"logps/chosen": -0.27874043583869934,
|
| 230 |
+
"logps/rejected": -0.3582325279712677,
|
| 231 |
+
"loss": 0.9163,
|
| 232 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 233 |
+
"rewards/chosen": -0.41811060905456543,
|
| 234 |
+
"rewards/margins": 0.11923813819885254,
|
| 235 |
+
"rewards/rejected": -0.537348747253418,
|
| 236 |
+
"step": 130
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.12433392539964476,
|
| 240 |
+
"grad_norm": 0.10971464216709137,
|
| 241 |
+
"learning_rate": 4.893298743830168e-06,
|
| 242 |
+
"logits/chosen": 14.18798828125,
|
| 243 |
+
"logits/rejected": 14.993026733398438,
|
| 244 |
+
"logps/chosen": -0.2750400900840759,
|
| 245 |
+
"logps/rejected": -0.39451608061790466,
|
| 246 |
+
"loss": 0.9098,
|
| 247 |
+
"rewards/accuracies": 0.574999988079071,
|
| 248 |
+
"rewards/chosen": -0.4125601351261139,
|
| 249 |
+
"rewards/margins": 0.17921395599842072,
|
| 250 |
+
"rewards/rejected": -0.5917741060256958,
|
| 251 |
+
"step": 140
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.13321492007104796,
|
| 255 |
+
"grad_norm": 0.09321591258049011,
|
| 256 |
+
"learning_rate": 4.8776412907378845e-06,
|
| 257 |
+
"logits/chosen": 12.775139808654785,
|
| 258 |
+
"logits/rejected": 13.751996994018555,
|
| 259 |
+
"logps/chosen": -0.28446996212005615,
|
| 260 |
+
"logps/rejected": -0.36404967308044434,
|
| 261 |
+
"loss": 0.9104,
|
| 262 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 263 |
+
"rewards/chosen": -0.42670494318008423,
|
| 264 |
+
"rewards/margins": 0.11936960369348526,
|
| 265 |
+
"rewards/rejected": -0.5460745096206665,
|
| 266 |
+
"step": 150
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.13321492007104796,
|
| 270 |
+
"eval_logits/chosen": 12.97266960144043,
|
| 271 |
+
"eval_logits/rejected": 13.47339916229248,
|
| 272 |
+
"eval_logps/chosen": -0.27297571301460266,
|
| 273 |
+
"eval_logps/rejected": -0.36854612827301025,
|
| 274 |
+
"eval_loss": 0.9143257737159729,
|
| 275 |
+
"eval_rewards/accuracies": 0.5824176073074341,
|
| 276 |
+
"eval_rewards/chosen": -0.4094635546207428,
|
| 277 |
+
"eval_rewards/margins": 0.14335563778877258,
|
| 278 |
+
"eval_rewards/rejected": -0.5528191924095154,
|
| 279 |
+
"eval_runtime": 25.2406,
|
| 280 |
+
"eval_samples_per_second": 28.842,
|
| 281 |
+
"eval_steps_per_second": 3.605,
|
| 282 |
+
"step": 150
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.14209591474245115,
|
| 286 |
+
"grad_norm": 0.11029861867427826,
|
| 287 |
+
"learning_rate": 4.860940925593703e-06,
|
| 288 |
+
"logits/chosen": 12.677947998046875,
|
| 289 |
+
"logits/rejected": 13.396716117858887,
|
| 290 |
+
"logps/chosen": -0.2631794512271881,
|
| 291 |
+
"logps/rejected": -0.37102141976356506,
|
| 292 |
+
"loss": 0.9051,
|
| 293 |
+
"rewards/accuracies": 0.550000011920929,
|
| 294 |
+
"rewards/chosen": -0.3947691321372986,
|
| 295 |
+
"rewards/margins": 0.161762997508049,
|
| 296 |
+
"rewards/rejected": -0.5565321445465088,
|
| 297 |
+
"step": 160
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.15097690941385436,
|
| 301 |
+
"grad_norm": 0.15728294849395752,
|
| 302 |
+
"learning_rate": 4.84320497372973e-06,
|
| 303 |
+
"logits/chosen": 12.620219230651855,
|
| 304 |
+
"logits/rejected": 13.189640998840332,
|
| 305 |
+
"logps/chosen": -0.2947639524936676,
|
| 306 |
+
"logps/rejected": -0.3843482732772827,
|
| 307 |
+
"loss": 0.8906,
|
| 308 |
+
"rewards/accuracies": 0.5625,
|
| 309 |
+
"rewards/chosen": -0.442145973443985,
|
| 310 |
+
"rewards/margins": 0.13437646627426147,
|
| 311 |
+
"rewards/rejected": -0.5765224099159241,
|
| 312 |
+
"step": 170
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.15985790408525755,
|
| 316 |
+
"grad_norm": 0.31504154205322266,
|
| 317 |
+
"learning_rate": 4.824441214720629e-06,
|
| 318 |
+
"logits/chosen": 11.487619400024414,
|
| 319 |
+
"logits/rejected": 12.33470344543457,
|
| 320 |
+
"logps/chosen": -0.271095871925354,
|
| 321 |
+
"logps/rejected": -0.4252637028694153,
|
| 322 |
+
"loss": 0.8766,
|
| 323 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 324 |
+
"rewards/chosen": -0.406643807888031,
|
| 325 |
+
"rewards/margins": 0.2312517911195755,
|
| 326 |
+
"rewards/rejected": -0.6378955245018005,
|
| 327 |
+
"step": 180
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.16873889875666073,
|
| 331 |
+
"grad_norm": 0.19222252070903778,
|
| 332 |
+
"learning_rate": 4.804657878971252e-06,
|
| 333 |
+
"logits/chosen": 10.093737602233887,
|
| 334 |
+
"logits/rejected": 10.851752281188965,
|
| 335 |
+
"logps/chosen": -0.2679918110370636,
|
| 336 |
+
"logps/rejected": -0.437336266040802,
|
| 337 |
+
"loss": 0.884,
|
| 338 |
+
"rewards/accuracies": 0.637499988079071,
|
| 339 |
+
"rewards/chosen": -0.4019877314567566,
|
| 340 |
+
"rewards/margins": 0.2540166974067688,
|
| 341 |
+
"rewards/rejected": -0.6560044288635254,
|
| 342 |
+
"step": 190
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.17761989342806395,
|
| 346 |
+
"grad_norm": 0.2275688648223877,
|
| 347 |
+
"learning_rate": 4.783863644106502e-06,
|
| 348 |
+
"logits/chosen": 9.483477592468262,
|
| 349 |
+
"logits/rejected": 10.106366157531738,
|
| 350 |
+
"logps/chosen": -0.2957404553890228,
|
| 351 |
+
"logps/rejected": -0.40739065408706665,
|
| 352 |
+
"loss": 0.8767,
|
| 353 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 354 |
+
"rewards/chosen": -0.4436107575893402,
|
| 355 |
+
"rewards/margins": 0.16747523844242096,
|
| 356 |
+
"rewards/rejected": -0.6110859513282776,
|
| 357 |
+
"step": 200
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.17761989342806395,
|
| 361 |
+
"eval_logits/chosen": 8.491498947143555,
|
| 362 |
+
"eval_logits/rejected": 8.999146461486816,
|
| 363 |
+
"eval_logps/chosen": -0.3135836124420166,
|
| 364 |
+
"eval_logps/rejected": -0.4829566180706024,
|
| 365 |
+
"eval_loss": 0.8664290904998779,
|
| 366 |
+
"eval_rewards/accuracies": 0.6263736486434937,
|
| 367 |
+
"eval_rewards/chosen": -0.4703753888607025,
|
| 368 |
+
"eval_rewards/margins": 0.2540595233440399,
|
| 369 |
+
"eval_rewards/rejected": -0.7244349122047424,
|
| 370 |
+
"eval_runtime": 25.2553,
|
| 371 |
+
"eval_samples_per_second": 28.826,
|
| 372 |
+
"eval_steps_per_second": 3.603,
|
| 373 |
+
"step": 200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.18650088809946713,
|
| 377 |
+
"grad_norm": 0.27885496616363525,
|
| 378 |
+
"learning_rate": 4.762067631165049e-06,
|
| 379 |
+
"logits/chosen": 7.234966278076172,
|
| 380 |
+
"logits/rejected": 8.313450813293457,
|
| 381 |
+
"logps/chosen": -0.29102542996406555,
|
| 382 |
+
"logps/rejected": -0.49241799116134644,
|
| 383 |
+
"loss": 0.8556,
|
| 384 |
+
"rewards/accuracies": 0.637499988079071,
|
| 385 |
+
"rewards/chosen": -0.43653813004493713,
|
| 386 |
+
"rewards/margins": 0.3020888566970825,
|
| 387 |
+
"rewards/rejected": -0.7386269569396973,
|
| 388 |
+
"step": 210
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.19538188277087035,
|
| 392 |
+
"grad_norm": 0.29907363653182983,
|
| 393 |
+
"learning_rate": 4.7392794005985324e-06,
|
| 394 |
+
"logits/chosen": 7.907521724700928,
|
| 395 |
+
"logits/rejected": 8.253190994262695,
|
| 396 |
+
"logps/chosen": -0.33691853284835815,
|
| 397 |
+
"logps/rejected": -0.4829257130622864,
|
| 398 |
+
"loss": 0.8236,
|
| 399 |
+
"rewards/accuracies": 0.550000011920929,
|
| 400 |
+
"rewards/chosen": -0.5053777694702148,
|
| 401 |
+
"rewards/margins": 0.21901080012321472,
|
| 402 |
+
"rewards/rejected": -0.724388599395752,
|
| 403 |
+
"step": 220
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.20426287744227353,
|
| 407 |
+
"grad_norm": 0.282474547624588,
|
| 408 |
+
"learning_rate": 4.715508948078037e-06,
|
| 409 |
+
"logits/chosen": 6.367492198944092,
|
| 410 |
+
"logits/rejected": 6.273728370666504,
|
| 411 |
+
"logps/chosen": -0.3519875705242157,
|
| 412 |
+
"logps/rejected": -0.5284813642501831,
|
| 413 |
+
"loss": 0.8027,
|
| 414 |
+
"rewards/accuracies": 0.625,
|
| 415 |
+
"rewards/chosen": -0.5279813408851624,
|
| 416 |
+
"rewards/margins": 0.2647407650947571,
|
| 417 |
+
"rewards/rejected": -0.7927221059799194,
|
| 418 |
+
"step": 230
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.21314387211367672,
|
| 422 |
+
"grad_norm": 0.327765554189682,
|
| 423 |
+
"learning_rate": 4.690766700109659e-06,
|
| 424 |
+
"logits/chosen": 5.090893268585205,
|
| 425 |
+
"logits/rejected": 4.768380165100098,
|
| 426 |
+
"logps/chosen": -0.3851698040962219,
|
| 427 |
+
"logps/rejected": -0.6464222073554993,
|
| 428 |
+
"loss": 0.7898,
|
| 429 |
+
"rewards/accuracies": 0.612500011920929,
|
| 430 |
+
"rewards/chosen": -0.5777546167373657,
|
| 431 |
+
"rewards/margins": 0.391878604888916,
|
| 432 |
+
"rewards/rejected": -0.9696332812309265,
|
| 433 |
+
"step": 240
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.22202486678507993,
|
| 437 |
+
"grad_norm": 0.4895865321159363,
|
| 438 |
+
"learning_rate": 4.665063509461098e-06,
|
| 439 |
+
"logits/chosen": 4.056812286376953,
|
| 440 |
+
"logits/rejected": 3.723601818084717,
|
| 441 |
+
"logps/chosen": -0.4400455951690674,
|
| 442 |
+
"logps/rejected": -0.7731422781944275,
|
| 443 |
+
"loss": 0.7626,
|
| 444 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 445 |
+
"rewards/chosen": -0.6600683927536011,
|
| 446 |
+
"rewards/margins": 0.49964505434036255,
|
| 447 |
+
"rewards/rejected": -1.1597135066986084,
|
| 448 |
+
"step": 250
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.22202486678507993,
|
| 452 |
+
"eval_logits/chosen": 2.420060396194458,
|
| 453 |
+
"eval_logits/rejected": 2.1626052856445312,
|
| 454 |
+
"eval_logps/chosen": -0.4724067151546478,
|
| 455 |
+
"eval_logps/rejected": -0.8418064117431641,
|
| 456 |
+
"eval_loss": 0.7631083130836487,
|
| 457 |
+
"eval_rewards/accuracies": 0.6483516693115234,
|
| 458 |
+
"eval_rewards/chosen": -0.7086100578308105,
|
| 459 |
+
"eval_rewards/margins": 0.5540997385978699,
|
| 460 |
+
"eval_rewards/rejected": -1.2627097368240356,
|
| 461 |
+
"eval_runtime": 25.2418,
|
| 462 |
+
"eval_samples_per_second": 28.841,
|
| 463 |
+
"eval_steps_per_second": 3.605,
|
| 464 |
+
"step": 250
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.23090586145648312,
|
| 468 |
+
"grad_norm": 0.46291017532348633,
|
| 469 |
+
"learning_rate": 4.638410650401267e-06,
|
| 470 |
+
"logits/chosen": 1.5297390222549438,
|
| 471 |
+
"logits/rejected": 1.1381648778915405,
|
| 472 |
+
"logps/chosen": -0.4418027997016907,
|
| 473 |
+
"logps/rejected": -1.0542564392089844,
|
| 474 |
+
"loss": 0.7026,
|
| 475 |
+
"rewards/accuracies": 0.762499988079071,
|
| 476 |
+
"rewards/chosen": -0.6627041697502136,
|
| 477 |
+
"rewards/margins": 0.9186803698539734,
|
| 478 |
+
"rewards/rejected": -1.5813844203948975,
|
| 479 |
+
"step": 260
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.23978685612788633,
|
| 483 |
+
"grad_norm": 0.9783313870429993,
|
| 484 |
+
"learning_rate": 4.610819813755038e-06,
|
| 485 |
+
"logits/chosen": 2.8311033248901367,
|
| 486 |
+
"logits/rejected": 1.9742711782455444,
|
| 487 |
+
"logps/chosen": -0.5430587530136108,
|
| 488 |
+
"logps/rejected": -0.9841039776802063,
|
| 489 |
+
"loss": 0.7317,
|
| 490 |
+
"rewards/accuracies": 0.699999988079071,
|
| 491 |
+
"rewards/chosen": -0.814588189125061,
|
| 492 |
+
"rewards/margins": 0.6615679860115051,
|
| 493 |
+
"rewards/rejected": -1.4761559963226318,
|
| 494 |
+
"step": 270
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.24866785079928952,
|
| 498 |
+
"grad_norm": 2.102562189102173,
|
| 499 |
+
"learning_rate": 4.582303101775249e-06,
|
| 500 |
+
"logits/chosen": 1.8241952657699585,
|
| 501 |
+
"logits/rejected": 0.8777934312820435,
|
| 502 |
+
"logps/chosen": -0.5624039769172668,
|
| 503 |
+
"logps/rejected": -1.1460126638412476,
|
| 504 |
+
"loss": 0.6887,
|
| 505 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 506 |
+
"rewards/chosen": -0.8436058163642883,
|
| 507 |
+
"rewards/margins": 0.8754131197929382,
|
| 508 |
+
"rewards/rejected": -1.7190189361572266,
|
| 509 |
+
"step": 280
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.25754884547069273,
|
| 513 |
+
"grad_norm": 0.9813026189804077,
|
| 514 |
+
"learning_rate": 4.55287302283426e-06,
|
| 515 |
+
"logits/chosen": 2.370732069015503,
|
| 516 |
+
"logits/rejected": 1.4697134494781494,
|
| 517 |
+
"logps/chosen": -0.6739786863327026,
|
| 518 |
+
"logps/rejected": -1.6581566333770752,
|
| 519 |
+
"loss": 0.5695,
|
| 520 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 521 |
+
"rewards/chosen": -1.0109679698944092,
|
| 522 |
+
"rewards/margins": 1.476266622543335,
|
| 523 |
+
"rewards/rejected": -2.487234592437744,
|
| 524 |
+
"step": 290
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.2664298401420959,
|
| 528 |
+
"grad_norm": 2.187314510345459,
|
| 529 |
+
"learning_rate": 4.522542485937369e-06,
|
| 530 |
+
"logits/chosen": 1.6230781078338623,
|
| 531 |
+
"logits/rejected": 0.5460122227668762,
|
| 532 |
+
"logps/chosen": -0.6433733701705933,
|
| 533 |
+
"logps/rejected": -2.1001811027526855,
|
| 534 |
+
"loss": 0.5366,
|
| 535 |
+
"rewards/accuracies": 0.762499988079071,
|
| 536 |
+
"rewards/chosen": -0.9650601148605347,
|
| 537 |
+
"rewards/margins": 2.185211658477783,
|
| 538 |
+
"rewards/rejected": -3.1502718925476074,
|
| 539 |
+
"step": 300
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.2664298401420959,
|
| 543 |
+
"eval_logits/chosen": 1.4087599515914917,
|
| 544 |
+
"eval_logits/rejected": 0.7888947129249573,
|
| 545 |
+
"eval_logps/chosen": -0.7579545974731445,
|
| 546 |
+
"eval_logps/rejected": -2.0049116611480713,
|
| 547 |
+
"eval_loss": 0.551510214805603,
|
| 548 |
+
"eval_rewards/accuracies": 0.6813187003135681,
|
| 549 |
+
"eval_rewards/chosen": -1.1369318962097168,
|
| 550 |
+
"eval_rewards/margins": 1.8704355955123901,
|
| 551 |
+
"eval_rewards/rejected": -3.0073673725128174,
|
| 552 |
+
"eval_runtime": 25.2647,
|
| 553 |
+
"eval_samples_per_second": 28.815,
|
| 554 |
+
"eval_steps_per_second": 3.602,
|
| 555 |
+
"step": 300
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2753108348134991,
|
| 559 |
+
"grad_norm": 0.7035408616065979,
|
| 560 |
+
"learning_rate": 4.491324795060491e-06,
|
| 561 |
+
"logits/chosen": 1.5831315517425537,
|
| 562 |
+
"logits/rejected": 0.46730250120162964,
|
| 563 |
+
"logps/chosen": -0.7262418866157532,
|
| 564 |
+
"logps/rejected": -2.1209158897399902,
|
| 565 |
+
"loss": 0.5524,
|
| 566 |
+
"rewards/accuracies": 0.675000011920929,
|
| 567 |
+
"rewards/chosen": -1.0893628597259521,
|
| 568 |
+
"rewards/margins": 2.092010974884033,
|
| 569 |
+
"rewards/rejected": -3.1813735961914062,
|
| 570 |
+
"step": 310
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.2841918294849023,
|
| 574 |
+
"grad_norm": 0.5678634643554688,
|
| 575 |
+
"learning_rate": 4.4592336433146e-06,
|
| 576 |
+
"logits/chosen": 1.265734076499939,
|
| 577 |
+
"logits/rejected": 0.7576489448547363,
|
| 578 |
+
"logps/chosen": -0.7938942313194275,
|
| 579 |
+
"logps/rejected": -2.3495612144470215,
|
| 580 |
+
"loss": 0.5233,
|
| 581 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 582 |
+
"rewards/chosen": -1.1908413171768188,
|
| 583 |
+
"rewards/margins": 2.333500385284424,
|
| 584 |
+
"rewards/rejected": -3.5243420600891113,
|
| 585 |
+
"step": 320
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.29307282415630553,
|
| 589 |
+
"grad_norm": 1.1373224258422852,
|
| 590 |
+
"learning_rate": 4.426283106939474e-06,
|
| 591 |
+
"logits/chosen": 2.977414846420288,
|
| 592 |
+
"logits/rejected": 2.1573710441589355,
|
| 593 |
+
"logps/chosen": -0.8513160943984985,
|
| 594 |
+
"logps/rejected": -2.4125566482543945,
|
| 595 |
+
"loss": 0.556,
|
| 596 |
+
"rewards/accuracies": 0.6875,
|
| 597 |
+
"rewards/chosen": -1.2769742012023926,
|
| 598 |
+
"rewards/margins": 2.341860771179199,
|
| 599 |
+
"rewards/rejected": -3.6188347339630127,
|
| 600 |
+
"step": 330
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.3019538188277087,
|
| 604 |
+
"grad_norm": 4.7876176834106445,
|
| 605 |
+
"learning_rate": 4.3924876391293915e-06,
|
| 606 |
+
"logits/chosen": 2.4026589393615723,
|
| 607 |
+
"logits/rejected": 1.207395315170288,
|
| 608 |
+
"logps/chosen": -0.8529679179191589,
|
| 609 |
+
"logps/rejected": -2.456879138946533,
|
| 610 |
+
"loss": 0.5592,
|
| 611 |
+
"rewards/accuracies": 0.637499988079071,
|
| 612 |
+
"rewards/chosen": -1.279451847076416,
|
| 613 |
+
"rewards/margins": 2.4058666229248047,
|
| 614 |
+
"rewards/rejected": -3.6853187084198,
|
| 615 |
+
"step": 340
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.3108348134991119,
|
| 619 |
+
"grad_norm": 0.5053763389587402,
|
| 620 |
+
"learning_rate": 4.357862063693486e-06,
|
| 621 |
+
"logits/chosen": 2.434265375137329,
|
| 622 |
+
"logits/rejected": 1.2504141330718994,
|
| 623 |
+
"logps/chosen": -0.9489291310310364,
|
| 624 |
+
"logps/rejected": -2.8521530628204346,
|
| 625 |
+
"loss": 0.4737,
|
| 626 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 627 |
+
"rewards/chosen": -1.423393726348877,
|
| 628 |
+
"rewards/margins": 2.8548355102539062,
|
| 629 |
+
"rewards/rejected": -4.278229236602783,
|
| 630 |
+
"step": 350
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.3108348134991119,
|
| 634 |
+
"eval_logits/chosen": 1.6632592678070068,
|
| 635 |
+
"eval_logits/rejected": 1.235045075416565,
|
| 636 |
+
"eval_logps/chosen": -1.0692518949508667,
|
| 637 |
+
"eval_logps/rejected": -2.7428486347198486,
|
| 638 |
+
"eval_loss": 0.5021397471427917,
|
| 639 |
+
"eval_rewards/accuracies": 0.692307710647583,
|
| 640 |
+
"eval_rewards/chosen": -1.6038777828216553,
|
| 641 |
+
"eval_rewards/margins": 2.510395050048828,
|
| 642 |
+
"eval_rewards/rejected": -4.1142730712890625,
|
| 643 |
+
"eval_runtime": 25.2582,
|
| 644 |
+
"eval_samples_per_second": 28.822,
|
| 645 |
+
"eval_steps_per_second": 3.603,
|
| 646 |
+
"step": 350
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.3197158081705151,
|
| 650 |
+
"grad_norm": 0.8040274381637573,
|
| 651 |
+
"learning_rate": 4.322421568553529e-06,
|
| 652 |
+
"logits/chosen": 1.187036395072937,
|
| 653 |
+
"logits/rejected": 0.4290788769721985,
|
| 654 |
+
"logps/chosen": -1.1015206575393677,
|
| 655 |
+
"logps/rejected": -2.919748544692993,
|
| 656 |
+
"loss": 0.489,
|
| 657 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 658 |
+
"rewards/chosen": -1.6522810459136963,
|
| 659 |
+
"rewards/margins": 2.727341651916504,
|
| 660 |
+
"rewards/rejected": -4.379622459411621,
|
| 661 |
+
"step": 360
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.3285968028419183,
|
| 665 |
+
"grad_norm": 0.9299562573432922,
|
| 666 |
+
"learning_rate": 4.286181699082008e-06,
|
| 667 |
+
"logits/chosen": 2.5852127075195312,
|
| 668 |
+
"logits/rejected": 2.0419259071350098,
|
| 669 |
+
"logps/chosen": -1.1498607397079468,
|
| 670 |
+
"logps/rejected": -3.0336194038391113,
|
| 671 |
+
"loss": 0.4812,
|
| 672 |
+
"rewards/accuracies": 0.675000011920929,
|
| 673 |
+
"rewards/chosen": -1.724791169166565,
|
| 674 |
+
"rewards/margins": 2.8256375789642334,
|
| 675 |
+
"rewards/rejected": -4.55042839050293,
|
| 676 |
+
"step": 370
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.33747779751332146,
|
| 680 |
+
"grad_norm": 1.7739671468734741,
|
| 681 |
+
"learning_rate": 4.249158351283414e-06,
|
| 682 |
+
"logits/chosen": 2.246245861053467,
|
| 683 |
+
"logits/rejected": 1.5551975965499878,
|
| 684 |
+
"logps/chosen": -1.254900574684143,
|
| 685 |
+
"logps/rejected": -3.206178665161133,
|
| 686 |
+
"loss": 0.4651,
|
| 687 |
+
"rewards/accuracies": 0.737500011920929,
|
| 688 |
+
"rewards/chosen": -1.8823509216308594,
|
| 689 |
+
"rewards/margins": 2.926917552947998,
|
| 690 |
+
"rewards/rejected": -4.809267997741699,
|
| 691 |
+
"step": 380
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.3463587921847247,
|
| 695 |
+
"grad_norm": 4.380665302276611,
|
| 696 |
+
"learning_rate": 4.211367764821722e-06,
|
| 697 |
+
"logits/chosen": 3.0754549503326416,
|
| 698 |
+
"logits/rejected": 2.622124433517456,
|
| 699 |
+
"logps/chosen": -1.9250037670135498,
|
| 700 |
+
"logps/rejected": -3.69482421875,
|
| 701 |
+
"loss": 0.4292,
|
| 702 |
+
"rewards/accuracies": 0.800000011920929,
|
| 703 |
+
"rewards/chosen": -2.8875060081481934,
|
| 704 |
+
"rewards/margins": 2.6547305583953857,
|
| 705 |
+
"rewards/rejected": -5.542236328125,
|
| 706 |
+
"step": 390
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.3552397868561279,
|
| 710 |
+
"grad_norm": 1.5087212324142456,
|
| 711 |
+
"learning_rate": 4.172826515897146e-06,
|
| 712 |
+
"logits/chosen": 2.2718021869659424,
|
| 713 |
+
"logits/rejected": 1.8861210346221924,
|
| 714 |
+
"logps/chosen": -2.4473955631256104,
|
| 715 |
+
"logps/rejected": -4.387387752532959,
|
| 716 |
+
"loss": 0.3902,
|
| 717 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 718 |
+
"rewards/chosen": -3.671093702316284,
|
| 719 |
+
"rewards/margins": 2.9099888801574707,
|
| 720 |
+
"rewards/rejected": -6.581082344055176,
|
| 721 |
+
"step": 400
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.3552397868561279,
|
| 725 |
+
"eval_logits/chosen": 1.759078860282898,
|
| 726 |
+
"eval_logits/rejected": 1.5246928930282593,
|
| 727 |
+
"eval_logps/chosen": -2.720665454864502,
|
| 728 |
+
"eval_logps/rejected": -4.613493919372559,
|
| 729 |
+
"eval_loss": 0.4054907560348511,
|
| 730 |
+
"eval_rewards/accuracies": 0.8791208863258362,
|
| 731 |
+
"eval_rewards/chosen": -4.080998420715332,
|
| 732 |
+
"eval_rewards/margins": 2.839242696762085,
|
| 733 |
+
"eval_rewards/rejected": -6.920241355895996,
|
| 734 |
+
"eval_runtime": 25.2363,
|
| 735 |
+
"eval_samples_per_second": 28.847,
|
| 736 |
+
"eval_steps_per_second": 3.606,
|
| 737 |
+
"step": 400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3641207815275311,
|
| 741 |
+
"grad_norm": 6.079421043395996,
|
| 742 |
+
"learning_rate": 4.133551509975264e-06,
|
| 743 |
+
"logits/chosen": 1.8841949701309204,
|
| 744 |
+
"logits/rejected": 1.3479797840118408,
|
| 745 |
+
"logps/chosen": -2.517265796661377,
|
| 746 |
+
"logps/rejected": -4.453648567199707,
|
| 747 |
+
"loss": 0.3977,
|
| 748 |
+
"rewards/accuracies": 0.862500011920929,
|
| 749 |
+
"rewards/chosen": -3.7758986949920654,
|
| 750 |
+
"rewards/margins": 2.9045748710632324,
|
| 751 |
+
"rewards/rejected": -6.680473327636719,
|
| 752 |
+
"step": 410
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.37300177619893427,
|
| 756 |
+
"grad_norm": 3.0998194217681885,
|
| 757 |
+
"learning_rate": 4.093559974371725e-06,
|
| 758 |
+
"logits/chosen": 1.6409276723861694,
|
| 759 |
+
"logits/rejected": 1.2141990661621094,
|
| 760 |
+
"logps/chosen": -2.2561168670654297,
|
| 761 |
+
"logps/rejected": -4.470211029052734,
|
| 762 |
+
"loss": 0.3527,
|
| 763 |
+
"rewards/accuracies": 0.875,
|
| 764 |
+
"rewards/chosen": -3.3841750621795654,
|
| 765 |
+
"rewards/margins": 3.321141004562378,
|
| 766 |
+
"rewards/rejected": -6.70531702041626,
|
| 767 |
+
"step": 420
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.38188277087033745,
|
| 771 |
+
"grad_norm": 6.982161045074463,
|
| 772 |
+
"learning_rate": 4.052869450695776e-06,
|
| 773 |
+
"logits/chosen": 2.835188388824463,
|
| 774 |
+
"logits/rejected": 2.3657329082489014,
|
| 775 |
+
"logps/chosen": -2.8557300567626953,
|
| 776 |
+
"logps/rejected": -5.075521469116211,
|
| 777 |
+
"loss": 0.387,
|
| 778 |
+
"rewards/accuracies": 0.887499988079071,
|
| 779 |
+
"rewards/chosen": -4.283595085144043,
|
| 780 |
+
"rewards/margins": 3.3296875953674316,
|
| 781 |
+
"rewards/rejected": -7.613282680511475,
|
| 782 |
+
"step": 430
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.3907637655417407,
|
| 786 |
+
"grad_norm": 2.139338970184326,
|
| 787 |
+
"learning_rate": 4.011497787155938e-06,
|
| 788 |
+
"logits/chosen": 2.126509189605713,
|
| 789 |
+
"logits/rejected": 1.459567904472351,
|
| 790 |
+
"logps/chosen": -3.1412863731384277,
|
| 791 |
+
"logps/rejected": -5.423466682434082,
|
| 792 |
+
"loss": 0.3611,
|
| 793 |
+
"rewards/accuracies": 0.925000011920929,
|
| 794 |
+
"rewards/chosen": -4.711928844451904,
|
| 795 |
+
"rewards/margins": 3.4232699871063232,
|
| 796 |
+
"rewards/rejected": -8.135198593139648,
|
| 797 |
+
"step": 440
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.3996447602131439,
|
| 801 |
+
"grad_norm": 1.7899377346038818,
|
| 802 |
+
"learning_rate": 3.969463130731183e-06,
|
| 803 |
+
"logits/chosen": 2.4551379680633545,
|
| 804 |
+
"logits/rejected": 2.0784289836883545,
|
| 805 |
+
"logps/chosen": -3.098043203353882,
|
| 806 |
+
"logps/rejected": -5.300747871398926,
|
| 807 |
+
"loss": 0.354,
|
| 808 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 809 |
+
"rewards/chosen": -4.647065162658691,
|
| 810 |
+
"rewards/margins": 3.3040566444396973,
|
| 811 |
+
"rewards/rejected": -7.951122283935547,
|
| 812 |
+
"step": 450
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.3996447602131439,
|
| 816 |
+
"eval_logits/chosen": 1.9761625528335571,
|
| 817 |
+
"eval_logits/rejected": 1.6654667854309082,
|
| 818 |
+
"eval_logps/chosen": -2.8789772987365723,
|
| 819 |
+
"eval_logps/rejected": -5.1105055809021,
|
| 820 |
+
"eval_loss": 0.36211252212524414,
|
| 821 |
+
"eval_rewards/accuracies": 0.8791208863258362,
|
| 822 |
+
"eval_rewards/chosen": -4.3184661865234375,
|
| 823 |
+
"eval_rewards/margins": 3.347292423248291,
|
| 824 |
+
"eval_rewards/rejected": -7.6657586097717285,
|
| 825 |
+
"eval_runtime": 25.2549,
|
| 826 |
+
"eval_samples_per_second": 28.826,
|
| 827 |
+
"eval_steps_per_second": 3.603,
|
| 828 |
+
"step": 450
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.40852575488454707,
|
| 832 |
+
"grad_norm": 1.9171936511993408,
|
| 833 |
+
"learning_rate": 3.92678391921108e-06,
|
| 834 |
+
"logits/chosen": 2.1672446727752686,
|
| 835 |
+
"logits/rejected": 1.6228408813476562,
|
| 836 |
+
"logps/chosen": -2.688931703567505,
|
| 837 |
+
"logps/rejected": -5.2408246994018555,
|
| 838 |
+
"loss": 0.3266,
|
| 839 |
+
"rewards/accuracies": 0.887499988079071,
|
| 840 |
+
"rewards/chosen": -4.033397197723389,
|
| 841 |
+
"rewards/margins": 3.8278393745422363,
|
| 842 |
+
"rewards/rejected": -7.861237525939941,
|
| 843 |
+
"step": 460
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.41740674955595025,
|
| 847 |
+
"grad_norm": 1.702635407447815,
|
| 848 |
+
"learning_rate": 3.88347887310836e-06,
|
| 849 |
+
"logits/chosen": 2.3164448738098145,
|
| 850 |
+
"logits/rejected": 2.047529697418213,
|
| 851 |
+
"logps/chosen": -2.6861701011657715,
|
| 852 |
+
"logps/rejected": -5.629918098449707,
|
| 853 |
+
"loss": 0.3339,
|
| 854 |
+
"rewards/accuracies": 0.925000011920929,
|
| 855 |
+
"rewards/chosen": -4.02925443649292,
|
| 856 |
+
"rewards/margins": 4.415622711181641,
|
| 857 |
+
"rewards/rejected": -8.444877624511719,
|
| 858 |
+
"step": 470
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.42628774422735344,
|
| 862 |
+
"grad_norm": 2.48634934425354,
|
| 863 |
+
"learning_rate": 3.839566987447492e-06,
|
| 864 |
+
"logits/chosen": 2.5225472450256348,
|
| 865 |
+
"logits/rejected": 2.0870003700256348,
|
| 866 |
+
"logps/chosen": -3.041111946105957,
|
| 867 |
+
"logps/rejected": -5.3499016761779785,
|
| 868 |
+
"loss": 0.3226,
|
| 869 |
+
"rewards/accuracies": 0.875,
|
| 870 |
+
"rewards/chosen": -4.561667442321777,
|
| 871 |
+
"rewards/margins": 3.4631850719451904,
|
| 872 |
+
"rewards/rejected": -8.024852752685547,
|
| 873 |
+
"step": 480
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.4351687388987567,
|
| 877 |
+
"grad_norm": 4.728499412536621,
|
| 878 |
+
"learning_rate": 3.795067523432826e-06,
|
| 879 |
+
"logits/chosen": 2.33893084526062,
|
| 880 |
+
"logits/rejected": 1.7909936904907227,
|
| 881 |
+
"logps/chosen": -2.7356209754943848,
|
| 882 |
+
"logps/rejected": -5.33417272567749,
|
| 883 |
+
"loss": 0.322,
|
| 884 |
+
"rewards/accuracies": 0.9375,
|
| 885 |
+
"rewards/chosen": -4.103431224822998,
|
| 886 |
+
"rewards/margins": 3.8978283405303955,
|
| 887 |
+
"rewards/rejected": -8.001258850097656,
|
| 888 |
+
"step": 490
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.44404973357015987,
|
| 892 |
+
"grad_norm": 8.412679672241211,
|
| 893 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 894 |
+
"logits/chosen": 2.788668632507324,
|
| 895 |
+
"logits/rejected": 2.439873695373535,
|
| 896 |
+
"logps/chosen": -3.3219153881073,
|
| 897 |
+
"logps/rejected": -5.992051124572754,
|
| 898 |
+
"loss": 0.3075,
|
| 899 |
+
"rewards/accuracies": 0.875,
|
| 900 |
+
"rewards/chosen": -4.98287296295166,
|
| 901 |
+
"rewards/margins": 4.005204200744629,
|
| 902 |
+
"rewards/rejected": -8.988077163696289,
|
| 903 |
+
"step": 500
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.44404973357015987,
|
| 907 |
+
"eval_logits/chosen": 2.165436029434204,
|
| 908 |
+
"eval_logits/rejected": 1.8186790943145752,
|
| 909 |
+
"eval_logps/chosen": -3.4299349784851074,
|
| 910 |
+
"eval_logps/rejected": -6.0660552978515625,
|
| 911 |
+
"eval_loss": 0.3319137990474701,
|
| 912 |
+
"eval_rewards/accuracies": 0.8901098966598511,
|
| 913 |
+
"eval_rewards/chosen": -5.14490270614624,
|
| 914 |
+
"eval_rewards/margins": 3.954181671142578,
|
| 915 |
+
"eval_rewards/rejected": -9.09908390045166,
|
| 916 |
+
"eval_runtime": 25.2602,
|
| 917 |
+
"eval_samples_per_second": 28.82,
|
| 918 |
+
"eval_steps_per_second": 3.603,
|
| 919 |
+
"step": 500
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.45293072824156305,
|
| 923 |
+
"grad_norm": 2.8339133262634277,
|
| 924 |
+
"learning_rate": 3.7043841852542884e-06,
|
| 925 |
+
"logits/chosen": 2.109018325805664,
|
| 926 |
+
"logits/rejected": 1.6996265649795532,
|
| 927 |
+
"logps/chosen": -3.288560390472412,
|
| 928 |
+
"logps/rejected": -5.986764430999756,
|
| 929 |
+
"loss": 0.3116,
|
| 930 |
+
"rewards/accuracies": 0.862500011920929,
|
| 931 |
+
"rewards/chosen": -4.932840824127197,
|
| 932 |
+
"rewards/margins": 4.047306060791016,
|
| 933 |
+
"rewards/rejected": -8.980146408081055,
|
| 934 |
+
"step": 510
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.46181172291296624,
|
| 938 |
+
"grad_norm": 3.8578269481658936,
|
| 939 |
+
"learning_rate": 3.658240087799655e-06,
|
| 940 |
+
"logits/chosen": 1.7659969329833984,
|
| 941 |
+
"logits/rejected": 1.4596515893936157,
|
| 942 |
+
"logps/chosen": -3.0301425457000732,
|
| 943 |
+
"logps/rejected": -6.252682209014893,
|
| 944 |
+
"loss": 0.3015,
|
| 945 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 946 |
+
"rewards/chosen": -4.5452141761779785,
|
| 947 |
+
"rewards/margins": 4.833809852600098,
|
| 948 |
+
"rewards/rejected": -9.379022598266602,
|
| 949 |
+
"step": 520
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.4706927175843694,
|
| 953 |
+
"grad_norm": 2.7795143127441406,
|
| 954 |
+
"learning_rate": 3.611587947962319e-06,
|
| 955 |
+
"logits/chosen": 2.472006320953369,
|
| 956 |
+
"logits/rejected": 1.9400993585586548,
|
| 957 |
+
"logps/chosen": -3.2479186058044434,
|
| 958 |
+
"logps/rejected": -6.475512504577637,
|
| 959 |
+
"loss": 0.3006,
|
| 960 |
+
"rewards/accuracies": 0.925000011920929,
|
| 961 |
+
"rewards/chosen": -4.871877193450928,
|
| 962 |
+
"rewards/margins": 4.841391086578369,
|
| 963 |
+
"rewards/rejected": -9.713269233703613,
|
| 964 |
+
"step": 530
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.47957371225577267,
|
| 968 |
+
"grad_norm": 3.5200746059417725,
|
| 969 |
+
"learning_rate": 3.564448228912682e-06,
|
| 970 |
+
"logits/chosen": 2.911531925201416,
|
| 971 |
+
"logits/rejected": 2.2947440147399902,
|
| 972 |
+
"logps/chosen": -3.387556791305542,
|
| 973 |
+
"logps/rejected": -6.559035301208496,
|
| 974 |
+
"loss": 0.2684,
|
| 975 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 976 |
+
"rewards/chosen": -5.081335544586182,
|
| 977 |
+
"rewards/margins": 4.757218360900879,
|
| 978 |
+
"rewards/rejected": -9.838552474975586,
|
| 979 |
+
"step": 540
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 0.48845470692717585,
|
| 983 |
+
"grad_norm": 2.368495225906372,
|
| 984 |
+
"learning_rate": 3.516841607689501e-06,
|
| 985 |
+
"logits/chosen": 1.276886224746704,
|
| 986 |
+
"logits/rejected": 1.3811718225479126,
|
| 987 |
+
"logps/chosen": -2.9689180850982666,
|
| 988 |
+
"logps/rejected": -6.619606971740723,
|
| 989 |
+
"loss": 0.3159,
|
| 990 |
+
"rewards/accuracies": 0.9375,
|
| 991 |
+
"rewards/chosen": -4.4533772468566895,
|
| 992 |
+
"rewards/margins": 5.4760332107543945,
|
| 993 |
+
"rewards/rejected": -9.929410934448242,
|
| 994 |
+
"step": 550
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.48845470692717585,
|
| 998 |
+
"eval_logits/chosen": 2.2735824584960938,
|
| 999 |
+
"eval_logits/rejected": 1.9788992404937744,
|
| 1000 |
+
"eval_logps/chosen": -3.698131799697876,
|
| 1001 |
+
"eval_logps/rejected": -6.64966344833374,
|
| 1002 |
+
"eval_loss": 0.30747923254966736,
|
| 1003 |
+
"eval_rewards/accuracies": 0.9230769276618958,
|
| 1004 |
+
"eval_rewards/chosen": -5.5471978187561035,
|
| 1005 |
+
"eval_rewards/margins": 4.427298545837402,
|
| 1006 |
+
"eval_rewards/rejected": -9.974496841430664,
|
| 1007 |
+
"eval_runtime": 25.2322,
|
| 1008 |
+
"eval_samples_per_second": 28.852,
|
| 1009 |
+
"eval_steps_per_second": 3.606,
|
| 1010 |
+
"step": 550
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.49733570159857904,
|
| 1014 |
+
"grad_norm": 1.901207685470581,
|
| 1015 |
+
"learning_rate": 3.4687889661302577e-06,
|
| 1016 |
+
"logits/chosen": 1.9734981060028076,
|
| 1017 |
+
"logits/rejected": 1.8617655038833618,
|
| 1018 |
+
"logps/chosen": -3.464953899383545,
|
| 1019 |
+
"logps/rejected": -6.746106147766113,
|
| 1020 |
+
"loss": 0.2912,
|
| 1021 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1022 |
+
"rewards/chosen": -5.1974310874938965,
|
| 1023 |
+
"rewards/margins": 4.921727180480957,
|
| 1024 |
+
"rewards/rejected": -10.119158744812012,
|
| 1025 |
+
"step": 560
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.5062166962699822,
|
| 1029 |
+
"grad_norm": 3.526299238204956,
|
| 1030 |
+
"learning_rate": 3.4203113817116955e-06,
|
| 1031 |
+
"logits/chosen": 3.0836069583892822,
|
| 1032 |
+
"logits/rejected": 2.75875186920166,
|
| 1033 |
+
"logps/chosen": -3.5981743335723877,
|
| 1034 |
+
"logps/rejected": -6.405775547027588,
|
| 1035 |
+
"loss": 0.3284,
|
| 1036 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1037 |
+
"rewards/chosen": -5.397261142730713,
|
| 1038 |
+
"rewards/margins": 4.211403846740723,
|
| 1039 |
+
"rewards/rejected": -9.608665466308594,
|
| 1040 |
+
"step": 570
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.5150976909413855,
|
| 1044 |
+
"grad_norm": 2.816272497177124,
|
| 1045 |
+
"learning_rate": 3.3714301183045382e-06,
|
| 1046 |
+
"logits/chosen": 3.0068678855895996,
|
| 1047 |
+
"logits/rejected": 2.466287136077881,
|
| 1048 |
+
"logps/chosen": -3.7227580547332764,
|
| 1049 |
+
"logps/rejected": -6.9311113357543945,
|
| 1050 |
+
"loss": 0.2728,
|
| 1051 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1052 |
+
"rewards/chosen": -5.584136962890625,
|
| 1053 |
+
"rewards/margins": 4.812530040740967,
|
| 1054 |
+
"rewards/rejected": -10.39666748046875,
|
| 1055 |
+
"step": 580
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.5239786856127886,
|
| 1059 |
+
"grad_norm": 2.433389902114868,
|
| 1060 |
+
"learning_rate": 3.3221666168464584e-06,
|
| 1061 |
+
"logits/chosen": 2.9992904663085938,
|
| 1062 |
+
"logits/rejected": 2.678699254989624,
|
| 1063 |
+
"logps/chosen": -3.540968418121338,
|
| 1064 |
+
"logps/rejected": -7.228091239929199,
|
| 1065 |
+
"loss": 0.2568,
|
| 1066 |
+
"rewards/accuracies": 0.9375,
|
| 1067 |
+
"rewards/chosen": -5.3114519119262695,
|
| 1068 |
+
"rewards/margins": 5.530684947967529,
|
| 1069 |
+
"rewards/rejected": -10.842137336730957,
|
| 1070 |
+
"step": 590
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.5328596802841918,
|
| 1074 |
+
"grad_norm": 2.7557125091552734,
|
| 1075 |
+
"learning_rate": 3.272542485937369e-06,
|
| 1076 |
+
"logits/chosen": 2.557410478591919,
|
| 1077 |
+
"logits/rejected": 2.331958770751953,
|
| 1078 |
+
"logps/chosen": -3.9404635429382324,
|
| 1079 |
+
"logps/rejected": -7.266766548156738,
|
| 1080 |
+
"loss": 0.2639,
|
| 1081 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1082 |
+
"rewards/chosen": -5.9106950759887695,
|
| 1083 |
+
"rewards/margins": 4.9894537925720215,
|
| 1084 |
+
"rewards/rejected": -10.900148391723633,
|
| 1085 |
+
"step": 600
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.5328596802841918,
|
| 1089 |
+
"eval_logits/chosen": 2.561415910720825,
|
| 1090 |
+
"eval_logits/rejected": 2.2971484661102295,
|
| 1091 |
+
"eval_logps/chosen": -4.015191555023193,
|
| 1092 |
+
"eval_logps/rejected": -7.222255229949951,
|
| 1093 |
+
"eval_loss": 0.28853774070739746,
|
| 1094 |
+
"eval_rewards/accuracies": 0.9340659379959106,
|
| 1095 |
+
"eval_rewards/chosen": -6.022787570953369,
|
| 1096 |
+
"eval_rewards/margins": 4.810595989227295,
|
| 1097 |
+
"eval_rewards/rejected": -10.833383560180664,
|
| 1098 |
+
"eval_runtime": 25.2577,
|
| 1099 |
+
"eval_samples_per_second": 28.823,
|
| 1100 |
+
"eval_steps_per_second": 3.603,
|
| 1101 |
+
"step": 600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.5417406749555951,
|
| 1105 |
+
"grad_norm": 2.582770824432373,
|
| 1106 |
+
"learning_rate": 3.222579492361179e-06,
|
| 1107 |
+
"logits/chosen": 2.7404181957244873,
|
| 1108 |
+
"logits/rejected": 2.540113687515259,
|
| 1109 |
+
"logps/chosen": -3.9154396057128906,
|
| 1110 |
+
"logps/rejected": -7.631985664367676,
|
| 1111 |
+
"loss": 0.2816,
|
| 1112 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1113 |
+
"rewards/chosen": -5.873159408569336,
|
| 1114 |
+
"rewards/margins": 5.574820041656494,
|
| 1115 |
+
"rewards/rejected": -11.447979927062988,
|
| 1116 |
+
"step": 610
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.5506216696269982,
|
| 1120 |
+
"grad_norm": 3.8167436122894287,
|
| 1121 |
+
"learning_rate": 3.1722995515381644e-06,
|
| 1122 |
+
"logits/chosen": 2.445218563079834,
|
| 1123 |
+
"logits/rejected": 2.288620710372925,
|
| 1124 |
+
"logps/chosen": -3.7501556873321533,
|
| 1125 |
+
"logps/rejected": -7.918539524078369,
|
| 1126 |
+
"loss": 0.2891,
|
| 1127 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1128 |
+
"rewards/chosen": -5.625233173370361,
|
| 1129 |
+
"rewards/margins": 6.252577304840088,
|
| 1130 |
+
"rewards/rejected": -11.877809524536133,
|
| 1131 |
+
"step": 620
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.5595026642984015,
|
| 1135 |
+
"grad_norm": 3.57536244392395,
|
| 1136 |
+
"learning_rate": 3.121724717912138e-06,
|
| 1137 |
+
"logits/chosen": 2.8337388038635254,
|
| 1138 |
+
"logits/rejected": 2.1241557598114014,
|
| 1139 |
+
"logps/chosen": -3.7040035724639893,
|
| 1140 |
+
"logps/rejected": -7.53197717666626,
|
| 1141 |
+
"loss": 0.2702,
|
| 1142 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1143 |
+
"rewards/chosen": -5.556005477905273,
|
| 1144 |
+
"rewards/margins": 5.741961479187012,
|
| 1145 |
+
"rewards/rejected": -11.297966003417969,
|
| 1146 |
+
"step": 630
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.5683836589698046,
|
| 1150 |
+
"grad_norm": 3.0520713329315186,
|
| 1151 |
+
"learning_rate": 3.0708771752766397e-06,
|
| 1152 |
+
"logits/chosen": 2.5255160331726074,
|
| 1153 |
+
"logits/rejected": 2.0742428302764893,
|
| 1154 |
+
"logps/chosen": -3.908573865890503,
|
| 1155 |
+
"logps/rejected": -7.60653829574585,
|
| 1156 |
+
"loss": 0.2598,
|
| 1157 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1158 |
+
"rewards/chosen": -5.862860679626465,
|
| 1159 |
+
"rewards/margins": 5.546946048736572,
|
| 1160 |
+
"rewards/rejected": -11.409807205200195,
|
| 1161 |
+
"step": 640
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.5772646536412078,
|
| 1165 |
+
"grad_norm": 8.067182540893555,
|
| 1166 |
+
"learning_rate": 3.019779227044398e-06,
|
| 1167 |
+
"logits/chosen": 1.6965067386627197,
|
| 1168 |
+
"logits/rejected": 1.644667625427246,
|
| 1169 |
+
"logps/chosen": -3.6588027477264404,
|
| 1170 |
+
"logps/rejected": -7.457572937011719,
|
| 1171 |
+
"loss": 0.2453,
|
| 1172 |
+
"rewards/accuracies": 0.9375,
|
| 1173 |
+
"rewards/chosen": -5.488204002380371,
|
| 1174 |
+
"rewards/margins": 5.698155403137207,
|
| 1175 |
+
"rewards/rejected": -11.186359405517578,
|
| 1176 |
+
"step": 650
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 0.5772646536412078,
|
| 1180 |
+
"eval_logits/chosen": 2.577754020690918,
|
| 1181 |
+
"eval_logits/rejected": 2.265626907348633,
|
| 1182 |
+
"eval_logps/chosen": -3.906606435775757,
|
| 1183 |
+
"eval_logps/rejected": -7.3099446296691895,
|
| 1184 |
+
"eval_loss": 0.27302286028862,
|
| 1185 |
+
"eval_rewards/accuracies": 0.9560439586639404,
|
| 1186 |
+
"eval_rewards/chosen": -5.859910011291504,
|
| 1187 |
+
"eval_rewards/margins": 5.105007171630859,
|
| 1188 |
+
"eval_rewards/rejected": -10.964917182922363,
|
| 1189 |
+
"eval_runtime": 25.2446,
|
| 1190 |
+
"eval_samples_per_second": 28.838,
|
| 1191 |
+
"eval_steps_per_second": 3.605,
|
| 1192 |
+
"step": 650
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.5861456483126111,
|
| 1196 |
+
"grad_norm": 2.9298255443573,
|
| 1197 |
+
"learning_rate": 2.9684532864643123e-06,
|
| 1198 |
+
"logits/chosen": 2.901702404022217,
|
| 1199 |
+
"logits/rejected": 2.595918655395508,
|
| 1200 |
+
"logps/chosen": -3.9949543476104736,
|
| 1201 |
+
"logps/rejected": -7.430356502532959,
|
| 1202 |
+
"loss": 0.2726,
|
| 1203 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 1204 |
+
"rewards/chosen": -5.992431163787842,
|
| 1205 |
+
"rewards/margins": 5.153104782104492,
|
| 1206 |
+
"rewards/rejected": -11.145535469055176,
|
| 1207 |
+
"step": 660
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.5950266429840142,
|
| 1211 |
+
"grad_norm": 3.085571050643921,
|
| 1212 |
+
"learning_rate": 2.9169218667902562e-06,
|
| 1213 |
+
"logits/chosen": 2.8471388816833496,
|
| 1214 |
+
"logits/rejected": 2.5500330924987793,
|
| 1215 |
+
"logps/chosen": -4.014147758483887,
|
| 1216 |
+
"logps/rejected": -7.298794746398926,
|
| 1217 |
+
"loss": 0.2484,
|
| 1218 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1219 |
+
"rewards/chosen": -6.02122163772583,
|
| 1220 |
+
"rewards/margins": 4.926970481872559,
|
| 1221 |
+
"rewards/rejected": -10.94819164276123,
|
| 1222 |
+
"step": 670
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.6039076376554174,
|
| 1226 |
+
"grad_norm": 2.3615477085113525,
|
| 1227 |
+
"learning_rate": 2.8652075714060296e-06,
|
| 1228 |
+
"logits/chosen": 2.495004177093506,
|
| 1229 |
+
"logits/rejected": 1.9873936176300049,
|
| 1230 |
+
"logps/chosen": -4.2625041007995605,
|
| 1231 |
+
"logps/rejected": -8.186586380004883,
|
| 1232 |
+
"loss": 0.2144,
|
| 1233 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1234 |
+
"rewards/chosen": -6.393756866455078,
|
| 1235 |
+
"rewards/margins": 5.886124610900879,
|
| 1236 |
+
"rewards/rejected": -12.279881477355957,
|
| 1237 |
+
"step": 680
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.6127886323268206,
|
| 1241 |
+
"grad_norm": 3.497316837310791,
|
| 1242 |
+
"learning_rate": 2.813333083910761e-06,
|
| 1243 |
+
"logits/chosen": 1.540621042251587,
|
| 1244 |
+
"logits/rejected": 1.2002273797988892,
|
| 1245 |
+
"logps/chosen": -3.71490740776062,
|
| 1246 |
+
"logps/rejected": -8.136199951171875,
|
| 1247 |
+
"loss": 0.2861,
|
| 1248 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1249 |
+
"rewards/chosen": -5.572361469268799,
|
| 1250 |
+
"rewards/margins": 6.631939888000488,
|
| 1251 |
+
"rewards/rejected": -12.204300880432129,
|
| 1252 |
+
"step": 690
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.6216696269982238,
|
| 1256 |
+
"grad_norm": 3.2540247440338135,
|
| 1257 |
+
"learning_rate": 2.761321158169134e-06,
|
| 1258 |
+
"logits/chosen": 2.776721477508545,
|
| 1259 |
+
"logits/rejected": 2.475888729095459,
|
| 1260 |
+
"logps/chosen": -4.3385329246521,
|
| 1261 |
+
"logps/rejected": -7.8828301429748535,
|
| 1262 |
+
"loss": 0.2885,
|
| 1263 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1264 |
+
"rewards/chosen": -6.507800102233887,
|
| 1265 |
+
"rewards/margins": 5.316445350646973,
|
| 1266 |
+
"rewards/rejected": -11.824244499206543,
|
| 1267 |
+
"step": 700
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 0.6216696269982238,
|
| 1271 |
+
"eval_logits/chosen": 2.4267516136169434,
|
| 1272 |
+
"eval_logits/rejected": 2.1245739459991455,
|
| 1273 |
+
"eval_logps/chosen": -3.9584598541259766,
|
| 1274 |
+
"eval_logps/rejected": -7.546706199645996,
|
| 1275 |
+
"eval_loss": 0.26512712240219116,
|
| 1276 |
+
"eval_rewards/accuracies": 0.9340659379959106,
|
| 1277 |
+
"eval_rewards/chosen": -5.937689781188965,
|
| 1278 |
+
"eval_rewards/margins": 5.382368564605713,
|
| 1279 |
+
"eval_rewards/rejected": -11.320058822631836,
|
| 1280 |
+
"eval_runtime": 25.2589,
|
| 1281 |
+
"eval_samples_per_second": 28.822,
|
| 1282 |
+
"eval_steps_per_second": 3.603,
|
| 1283 |
+
"step": 700
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.6305506216696269,
|
| 1287 |
+
"grad_norm": 2.582273483276367,
|
| 1288 |
+
"learning_rate": 2.70919460833079e-06,
|
| 1289 |
+
"logits/chosen": 2.2102553844451904,
|
| 1290 |
+
"logits/rejected": 1.9504516124725342,
|
| 1291 |
+
"logps/chosen": -3.7470879554748535,
|
| 1292 |
+
"logps/rejected": -7.761659145355225,
|
| 1293 |
+
"loss": 0.2078,
|
| 1294 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1295 |
+
"rewards/chosen": -5.620632171630859,
|
| 1296 |
+
"rewards/margins": 6.02185583114624,
|
| 1297 |
+
"rewards/rejected": -11.642488479614258,
|
| 1298 |
+
"step": 710
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.6394316163410302,
|
| 1302 |
+
"grad_norm": 2.2855663299560547,
|
| 1303 |
+
"learning_rate": 2.6569762988232838e-06,
|
| 1304 |
+
"logits/chosen": 2.4659409523010254,
|
| 1305 |
+
"logits/rejected": 2.0434811115264893,
|
| 1306 |
+
"logps/chosen": -3.4346442222595215,
|
| 1307 |
+
"logps/rejected": -7.470824241638184,
|
| 1308 |
+
"loss": 0.2346,
|
| 1309 |
+
"rewards/accuracies": 0.9375,
|
| 1310 |
+
"rewards/chosen": -5.151965618133545,
|
| 1311 |
+
"rewards/margins": 6.054270267486572,
|
| 1312 |
+
"rewards/rejected": -11.206236839294434,
|
| 1313 |
+
"step": 720
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.6483126110124334,
|
| 1317 |
+
"grad_norm": 2.038733959197998,
|
| 1318 |
+
"learning_rate": 2.604689134322999e-06,
|
| 1319 |
+
"logits/chosen": 2.270310878753662,
|
| 1320 |
+
"logits/rejected": 1.9651508331298828,
|
| 1321 |
+
"logps/chosen": -3.721379518508911,
|
| 1322 |
+
"logps/rejected": -7.776650905609131,
|
| 1323 |
+
"loss": 0.2354,
|
| 1324 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1325 |
+
"rewards/chosen": -5.582068920135498,
|
| 1326 |
+
"rewards/margins": 6.0829057693481445,
|
| 1327 |
+
"rewards/rejected": -11.6649751663208,
|
| 1328 |
+
"step": 730
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.6571936056838366,
|
| 1332 |
+
"grad_norm": 2.948915481567383,
|
| 1333 |
+
"learning_rate": 2.5523560497083927e-06,
|
| 1334 |
+
"logits/chosen": 2.358057737350464,
|
| 1335 |
+
"logits/rejected": 2.063586711883545,
|
| 1336 |
+
"logps/chosen": -3.6883864402770996,
|
| 1337 |
+
"logps/rejected": -7.352984428405762,
|
| 1338 |
+
"loss": 0.231,
|
| 1339 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1340 |
+
"rewards/chosen": -5.5325798988342285,
|
| 1341 |
+
"rewards/margins": 5.496896266937256,
|
| 1342 |
+
"rewards/rejected": -11.0294771194458,
|
| 1343 |
+
"step": 740
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.6660746003552398,
|
| 1347 |
+
"grad_norm": 3.661870002746582,
|
| 1348 |
+
"learning_rate": 2.5e-06,
|
| 1349 |
+
"logits/chosen": 2.6712796688079834,
|
| 1350 |
+
"logits/rejected": 2.393817901611328,
|
| 1351 |
+
"logps/chosen": -4.065141201019287,
|
| 1352 |
+
"logps/rejected": -7.9232306480407715,
|
| 1353 |
+
"loss": 0.2332,
|
| 1354 |
+
"rewards/accuracies": 0.875,
|
| 1355 |
+
"rewards/chosen": -6.097712516784668,
|
| 1356 |
+
"rewards/margins": 5.787134170532227,
|
| 1357 |
+
"rewards/rejected": -11.884846687316895,
|
| 1358 |
+
"step": 750
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.6660746003552398,
|
| 1362 |
+
"eval_logits/chosen": 2.4871790409088135,
|
| 1363 |
+
"eval_logits/rejected": 2.2864174842834473,
|
| 1364 |
+
"eval_logps/chosen": -4.176412105560303,
|
| 1365 |
+
"eval_logps/rejected": -8.009041786193848,
|
| 1366 |
+
"eval_loss": 0.2521994411945343,
|
| 1367 |
+
"eval_rewards/accuracies": 0.9340659379959106,
|
| 1368 |
+
"eval_rewards/chosen": -6.264617443084717,
|
| 1369 |
+
"eval_rewards/margins": 5.7489447593688965,
|
| 1370 |
+
"eval_rewards/rejected": -12.01356315612793,
|
| 1371 |
+
"eval_runtime": 25.2624,
|
| 1372 |
+
"eval_samples_per_second": 28.818,
|
| 1373 |
+
"eval_steps_per_second": 3.602,
|
| 1374 |
+
"step": 750
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.6749555950266429,
|
| 1378 |
+
"grad_norm": 1.992043375968933,
|
| 1379 |
+
"learning_rate": 2.447643950291608e-06,
|
| 1380 |
+
"logits/chosen": 1.8225914239883423,
|
| 1381 |
+
"logits/rejected": 1.842559814453125,
|
| 1382 |
+
"logps/chosen": -4.158253192901611,
|
| 1383 |
+
"logps/rejected": -8.65810775756836,
|
| 1384 |
+
"loss": 0.2104,
|
| 1385 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1386 |
+
"rewards/chosen": -6.237379550933838,
|
| 1387 |
+
"rewards/margins": 6.749783515930176,
|
| 1388 |
+
"rewards/rejected": -12.987161636352539,
|
| 1389 |
+
"step": 760
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.6838365896980462,
|
| 1393 |
+
"grad_norm": 1.9012709856033325,
|
| 1394 |
+
"learning_rate": 2.3953108656770018e-06,
|
| 1395 |
+
"logits/chosen": 2.491041660308838,
|
| 1396 |
+
"logits/rejected": 1.9490203857421875,
|
| 1397 |
+
"logps/chosen": -4.141987323760986,
|
| 1398 |
+
"logps/rejected": -8.790396690368652,
|
| 1399 |
+
"loss": 0.2236,
|
| 1400 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1401 |
+
"rewards/chosen": -6.212981224060059,
|
| 1402 |
+
"rewards/margins": 6.9726152420043945,
|
| 1403 |
+
"rewards/rejected": -13.185595512390137,
|
| 1404 |
+
"step": 770
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.6927175843694494,
|
| 1408 |
+
"grad_norm": 3.4377241134643555,
|
| 1409 |
+
"learning_rate": 2.3430237011767166e-06,
|
| 1410 |
+
"logits/chosen": 2.54376220703125,
|
| 1411 |
+
"logits/rejected": 2.2269883155822754,
|
| 1412 |
+
"logps/chosen": -3.7832961082458496,
|
| 1413 |
+
"logps/rejected": -8.058969497680664,
|
| 1414 |
+
"loss": 0.2449,
|
| 1415 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1416 |
+
"rewards/chosen": -5.674944877624512,
|
| 1417 |
+
"rewards/margins": 6.413510322570801,
|
| 1418 |
+
"rewards/rejected": -12.08845329284668,
|
| 1419 |
+
"step": 780
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.7015985790408525,
|
| 1423 |
+
"grad_norm": 1.365478277206421,
|
| 1424 |
+
"learning_rate": 2.290805391669212e-06,
|
| 1425 |
+
"logits/chosen": 3.386944532394409,
|
| 1426 |
+
"logits/rejected": 2.942134141921997,
|
| 1427 |
+
"logps/chosen": -3.8051178455352783,
|
| 1428 |
+
"logps/rejected": -8.201702117919922,
|
| 1429 |
+
"loss": 0.1764,
|
| 1430 |
+
"rewards/accuracies": 1.0,
|
| 1431 |
+
"rewards/chosen": -5.707675933837891,
|
| 1432 |
+
"rewards/margins": 6.594876289367676,
|
| 1433 |
+
"rewards/rejected": -12.302552223205566,
|
| 1434 |
+
"step": 790
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.7104795737122558,
|
| 1438 |
+
"grad_norm": 3.3714189529418945,
|
| 1439 |
+
"learning_rate": 2.238678841830867e-06,
|
| 1440 |
+
"logits/chosen": 2.690781593322754,
|
| 1441 |
+
"logits/rejected": 2.2246782779693604,
|
| 1442 |
+
"logps/chosen": -4.251287937164307,
|
| 1443 |
+
"logps/rejected": -8.268930435180664,
|
| 1444 |
+
"loss": 0.2429,
|
| 1445 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1446 |
+
"rewards/chosen": -6.376931667327881,
|
| 1447 |
+
"rewards/margins": 6.026463031768799,
|
| 1448 |
+
"rewards/rejected": -12.40339469909668,
|
| 1449 |
+
"step": 800
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.7104795737122558,
|
| 1453 |
+
"eval_logits/chosen": 2.453395128250122,
|
| 1454 |
+
"eval_logits/rejected": 2.2390329837799072,
|
| 1455 |
+
"eval_logps/chosen": -4.038145065307617,
|
| 1456 |
+
"eval_logps/rejected": -8.012212753295898,
|
| 1457 |
+
"eval_loss": 0.2468602955341339,
|
| 1458 |
+
"eval_rewards/accuracies": 0.9340659379959106,
|
| 1459 |
+
"eval_rewards/chosen": -6.057217597961426,
|
| 1460 |
+
"eval_rewards/margins": 5.961101055145264,
|
| 1461 |
+
"eval_rewards/rejected": -12.018318176269531,
|
| 1462 |
+
"eval_runtime": 25.2622,
|
| 1463 |
+
"eval_samples_per_second": 28.818,
|
| 1464 |
+
"eval_steps_per_second": 3.602,
|
| 1465 |
+
"step": 800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.7193605683836589,
|
| 1469 |
+
"grad_norm": 4.527968883514404,
|
| 1470 |
+
"learning_rate": 2.186666916089239e-06,
|
| 1471 |
+
"logits/chosen": 2.250300168991089,
|
| 1472 |
+
"logits/rejected": 2.254126787185669,
|
| 1473 |
+
"logps/chosen": -4.0314741134643555,
|
| 1474 |
+
"logps/rejected": -7.800264835357666,
|
| 1475 |
+
"loss": 0.2277,
|
| 1476 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1477 |
+
"rewards/chosen": -6.047211647033691,
|
| 1478 |
+
"rewards/margins": 5.65318489074707,
|
| 1479 |
+
"rewards/rejected": -11.700395584106445,
|
| 1480 |
+
"step": 810
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.7282415630550622,
|
| 1484 |
+
"grad_norm": 3.3826537132263184,
|
| 1485 |
+
"learning_rate": 2.134792428593971e-06,
|
| 1486 |
+
"logits/chosen": 2.260606050491333,
|
| 1487 |
+
"logits/rejected": 1.8174827098846436,
|
| 1488 |
+
"logps/chosen": -3.6692988872528076,
|
| 1489 |
+
"logps/rejected": -7.9340972900390625,
|
| 1490 |
+
"loss": 0.2428,
|
| 1491 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1492 |
+
"rewards/chosen": -5.503948211669922,
|
| 1493 |
+
"rewards/margins": 6.397197723388672,
|
| 1494 |
+
"rewards/rejected": -11.901147842407227,
|
| 1495 |
+
"step": 820
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.7371225577264654,
|
| 1499 |
+
"grad_norm": 5.055661678314209,
|
| 1500 |
+
"learning_rate": 2.0830781332097446e-06,
|
| 1501 |
+
"logits/chosen": 2.4391427040100098,
|
| 1502 |
+
"logits/rejected": 2.0999953746795654,
|
| 1503 |
+
"logps/chosen": -3.677706241607666,
|
| 1504 |
+
"logps/rejected": -7.840268135070801,
|
| 1505 |
+
"loss": 0.2288,
|
| 1506 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1507 |
+
"rewards/chosen": -5.516559600830078,
|
| 1508 |
+
"rewards/margins": 6.243842124938965,
|
| 1509 |
+
"rewards/rejected": -11.760400772094727,
|
| 1510 |
+
"step": 830
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.7460035523978685,
|
| 1514 |
+
"grad_norm": 9.756246566772461,
|
| 1515 |
+
"learning_rate": 2.031546713535688e-06,
|
| 1516 |
+
"logits/chosen": 2.5405280590057373,
|
| 1517 |
+
"logits/rejected": 2.3721981048583984,
|
| 1518 |
+
"logps/chosen": -3.955289840698242,
|
| 1519 |
+
"logps/rejected": -7.609635353088379,
|
| 1520 |
+
"loss": 0.2384,
|
| 1521 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1522 |
+
"rewards/chosen": -5.932934761047363,
|
| 1523 |
+
"rewards/margins": 5.481517314910889,
|
| 1524 |
+
"rewards/rejected": -11.41445255279541,
|
| 1525 |
+
"step": 840
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.7548845470692718,
|
| 1529 |
+
"grad_norm": 3.4563076496124268,
|
| 1530 |
+
"learning_rate": 1.9802207729556023e-06,
|
| 1531 |
+
"logits/chosen": 3.1546592712402344,
|
| 1532 |
+
"logits/rejected": 2.6494877338409424,
|
| 1533 |
+
"logps/chosen": -3.9850029945373535,
|
| 1534 |
+
"logps/rejected": -8.197628021240234,
|
| 1535 |
+
"loss": 0.2229,
|
| 1536 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1537 |
+
"rewards/chosen": -5.977504253387451,
|
| 1538 |
+
"rewards/margins": 6.3189377784729,
|
| 1539 |
+
"rewards/rejected": -12.296442031860352,
|
| 1540 |
+
"step": 850
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.7548845470692718,
|
| 1544 |
+
"eval_logits/chosen": 2.536773443222046,
|
| 1545 |
+
"eval_logits/rejected": 2.2915148735046387,
|
| 1546 |
+
"eval_logps/chosen": -4.12998628616333,
|
| 1547 |
+
"eval_logps/rejected": -8.214204788208008,
|
| 1548 |
+
"eval_loss": 0.2361687868833542,
|
| 1549 |
+
"eval_rewards/accuracies": 0.9340659379959106,
|
| 1550 |
+
"eval_rewards/chosen": -6.194979190826416,
|
| 1551 |
+
"eval_rewards/margins": 6.126327991485596,
|
| 1552 |
+
"eval_rewards/rejected": -12.321306228637695,
|
| 1553 |
+
"eval_runtime": 25.2618,
|
| 1554 |
+
"eval_samples_per_second": 28.818,
|
| 1555 |
+
"eval_steps_per_second": 3.602,
|
| 1556 |
+
"step": 850
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.7637655417406749,
|
| 1560 |
+
"grad_norm": 2.1946489810943604,
|
| 1561 |
+
"learning_rate": 1.9291228247233607e-06,
|
| 1562 |
+
"logits/chosen": 1.6681854724884033,
|
| 1563 |
+
"logits/rejected": 1.4272549152374268,
|
| 1564 |
+
"logps/chosen": -3.6183536052703857,
|
| 1565 |
+
"logps/rejected": -8.374089241027832,
|
| 1566 |
+
"loss": 0.2142,
|
| 1567 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1568 |
+
"rewards/chosen": -5.427530288696289,
|
| 1569 |
+
"rewards/margins": 7.133603572845459,
|
| 1570 |
+
"rewards/rejected": -12.56113338470459,
|
| 1571 |
+
"step": 860
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.7726465364120781,
|
| 1575 |
+
"grad_norm": 3.251861095428467,
|
| 1576 |
+
"learning_rate": 1.8782752820878636e-06,
|
| 1577 |
+
"logits/chosen": 3.6365694999694824,
|
| 1578 |
+
"logits/rejected": 3.002140522003174,
|
| 1579 |
+
"logps/chosen": -4.218847274780273,
|
| 1580 |
+
"logps/rejected": -8.535767555236816,
|
| 1581 |
+
"loss": 0.1927,
|
| 1582 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1583 |
+
"rewards/chosen": -6.328269958496094,
|
| 1584 |
+
"rewards/margins": 6.475382328033447,
|
| 1585 |
+
"rewards/rejected": -12.8036527633667,
|
| 1586 |
+
"step": 870
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.7815275310834814,
|
| 1590 |
+
"grad_norm": 2.358853816986084,
|
| 1591 |
+
"learning_rate": 1.827700448461836e-06,
|
| 1592 |
+
"logits/chosen": 2.647890567779541,
|
| 1593 |
+
"logits/rejected": 2.1392157077789307,
|
| 1594 |
+
"logps/chosen": -4.547138214111328,
|
| 1595 |
+
"logps/rejected": -9.499224662780762,
|
| 1596 |
+
"loss": 0.1945,
|
| 1597 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1598 |
+
"rewards/chosen": -6.820706844329834,
|
| 1599 |
+
"rewards/margins": 7.42812967300415,
|
| 1600 |
+
"rewards/rejected": -14.2488374710083,
|
| 1601 |
+
"step": 880
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.7904085257548845,
|
| 1605 |
+
"grad_norm": 2.04150390625,
|
| 1606 |
+
"learning_rate": 1.7774205076388207e-06,
|
| 1607 |
+
"logits/chosen": 2.7666563987731934,
|
| 1608 |
+
"logits/rejected": 2.417945623397827,
|
| 1609 |
+
"logps/chosen": -4.042238235473633,
|
| 1610 |
+
"logps/rejected": -7.750130653381348,
|
| 1611 |
+
"loss": 0.2228,
|
| 1612 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1613 |
+
"rewards/chosen": -6.063356876373291,
|
| 1614 |
+
"rewards/margins": 5.561838626861572,
|
| 1615 |
+
"rewards/rejected": -11.625195503234863,
|
| 1616 |
+
"step": 890
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.7992895204262878,
|
| 1620 |
+
"grad_norm": 3.5930328369140625,
|
| 1621 |
+
"learning_rate": 1.7274575140626318e-06,
|
| 1622 |
+
"logits/chosen": 2.7729713916778564,
|
| 1623 |
+
"logits/rejected": 2.426771640777588,
|
| 1624 |
+
"logps/chosen": -4.228175163269043,
|
| 1625 |
+
"logps/rejected": -8.267873764038086,
|
| 1626 |
+
"loss": 0.2134,
|
| 1627 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1628 |
+
"rewards/chosen": -6.342263221740723,
|
| 1629 |
+
"rewards/margins": 6.05954647064209,
|
| 1630 |
+
"rewards/rejected": -12.401810646057129,
|
| 1631 |
+
"step": 900
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.7992895204262878,
|
| 1635 |
+
"eval_logits/chosen": 2.5573580265045166,
|
| 1636 |
+
"eval_logits/rejected": 2.317056179046631,
|
| 1637 |
+
"eval_logps/chosen": -4.389712333679199,
|
| 1638 |
+
"eval_logps/rejected": -8.52020263671875,
|
| 1639 |
+
"eval_loss": 0.2303350269794464,
|
| 1640 |
+
"eval_rewards/accuracies": 0.9450549483299255,
|
| 1641 |
+
"eval_rewards/chosen": -6.584568023681641,
|
| 1642 |
+
"eval_rewards/margins": 6.195735931396484,
|
| 1643 |
+
"eval_rewards/rejected": -12.780303001403809,
|
| 1644 |
+
"eval_runtime": 25.2651,
|
| 1645 |
+
"eval_samples_per_second": 28.814,
|
| 1646 |
+
"eval_steps_per_second": 3.602,
|
| 1647 |
+
"step": 900
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.8081705150976909,
|
| 1651 |
+
"grad_norm": 3.19136118888855,
|
| 1652 |
+
"learning_rate": 1.677833383153542e-06,
|
| 1653 |
+
"logits/chosen": 2.2394022941589355,
|
| 1654 |
+
"logits/rejected": 2.2709288597106934,
|
| 1655 |
+
"logps/chosen": -4.338564872741699,
|
| 1656 |
+
"logps/rejected": -8.621049880981445,
|
| 1657 |
+
"loss": 0.2029,
|
| 1658 |
+
"rewards/accuracies": 0.9375,
|
| 1659 |
+
"rewards/chosen": -6.507847785949707,
|
| 1660 |
+
"rewards/margins": 6.423727512359619,
|
| 1661 |
+
"rewards/rejected": -12.931573867797852,
|
| 1662 |
+
"step": 910
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.8170515097690941,
|
| 1666 |
+
"grad_norm": 2.785442590713501,
|
| 1667 |
+
"learning_rate": 1.6285698816954626e-06,
|
| 1668 |
+
"logits/chosen": 2.669541835784912,
|
| 1669 |
+
"logits/rejected": 2.38188099861145,
|
| 1670 |
+
"logps/chosen": -4.103513717651367,
|
| 1671 |
+
"logps/rejected": -8.361233711242676,
|
| 1672 |
+
"loss": 0.2054,
|
| 1673 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1674 |
+
"rewards/chosen": -6.155270576477051,
|
| 1675 |
+
"rewards/margins": 6.386577606201172,
|
| 1676 |
+
"rewards/rejected": -12.541848182678223,
|
| 1677 |
+
"step": 920
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.8259325044404974,
|
| 1681 |
+
"grad_norm": 2.618563175201416,
|
| 1682 |
+
"learning_rate": 1.5796886182883053e-06,
|
| 1683 |
+
"logits/chosen": 2.5333399772644043,
|
| 1684 |
+
"logits/rejected": 2.3547558784484863,
|
| 1685 |
+
"logps/chosen": -3.701815128326416,
|
| 1686 |
+
"logps/rejected": -8.315168380737305,
|
| 1687 |
+
"loss": 0.1958,
|
| 1688 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1689 |
+
"rewards/chosen": -5.552721977233887,
|
| 1690 |
+
"rewards/margins": 6.9200310707092285,
|
| 1691 |
+
"rewards/rejected": -12.472753524780273,
|
| 1692 |
+
"step": 930
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 0.8348134991119005,
|
| 1696 |
+
"grad_norm": 2.640681266784668,
|
| 1697 |
+
"learning_rate": 1.5312110338697427e-06,
|
| 1698 |
+
"logits/chosen": 2.6260218620300293,
|
| 1699 |
+
"logits/rejected": 2.625610589981079,
|
| 1700 |
+
"logps/chosen": -3.761087417602539,
|
| 1701 |
+
"logps/rejected": -7.872547149658203,
|
| 1702 |
+
"loss": 0.2147,
|
| 1703 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1704 |
+
"rewards/chosen": -5.64163064956665,
|
| 1705 |
+
"rewards/margins": 6.167189598083496,
|
| 1706 |
+
"rewards/rejected": -11.808821678161621,
|
| 1707 |
+
"step": 940
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.8436944937833037,
|
| 1711 |
+
"grad_norm": 2.4091708660125732,
|
| 1712 |
+
"learning_rate": 1.4831583923105e-06,
|
| 1713 |
+
"logits/chosen": 2.63578462600708,
|
| 1714 |
+
"logits/rejected": 2.2209556102752686,
|
| 1715 |
+
"logps/chosen": -4.115639686584473,
|
| 1716 |
+
"logps/rejected": -8.535353660583496,
|
| 1717 |
+
"loss": 0.1832,
|
| 1718 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1719 |
+
"rewards/chosen": -6.173459529876709,
|
| 1720 |
+
"rewards/margins": 6.629571437835693,
|
| 1721 |
+
"rewards/rejected": -12.803030014038086,
|
| 1722 |
+
"step": 950
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 0.8436944937833037,
|
| 1726 |
+
"eval_logits/chosen": 2.6680469512939453,
|
| 1727 |
+
"eval_logits/rejected": 2.4164178371429443,
|
| 1728 |
+
"eval_logps/chosen": -4.110969066619873,
|
| 1729 |
+
"eval_logps/rejected": -8.377038955688477,
|
| 1730 |
+
"eval_loss": 0.22471140325069427,
|
| 1731 |
+
"eval_rewards/accuracies": 0.9560439586639404,
|
| 1732 |
+
"eval_rewards/chosen": -6.166452884674072,
|
| 1733 |
+
"eval_rewards/margins": 6.399104118347168,
|
| 1734 |
+
"eval_rewards/rejected": -12.565558433532715,
|
| 1735 |
+
"eval_runtime": 25.2357,
|
| 1736 |
+
"eval_samples_per_second": 28.848,
|
| 1737 |
+
"eval_steps_per_second": 3.606,
|
| 1738 |
+
"step": 950
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.8525754884547069,
|
| 1742 |
+
"grad_norm": 5.924206733703613,
|
| 1743 |
+
"learning_rate": 1.4355517710873184e-06,
|
| 1744 |
+
"logits/chosen": 1.8501287698745728,
|
| 1745 |
+
"logits/rejected": 1.9165807962417603,
|
| 1746 |
+
"logps/chosen": -3.9117112159729004,
|
| 1747 |
+
"logps/rejected": -8.62653923034668,
|
| 1748 |
+
"loss": 0.2384,
|
| 1749 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1750 |
+
"rewards/chosen": -5.86756706237793,
|
| 1751 |
+
"rewards/margins": 7.072243690490723,
|
| 1752 |
+
"rewards/rejected": -12.939809799194336,
|
| 1753 |
+
"step": 960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.8614564831261101,
|
| 1757 |
+
"grad_norm": 9.17930793762207,
|
| 1758 |
+
"learning_rate": 1.388412052037682e-06,
|
| 1759 |
+
"logits/chosen": 3.1926915645599365,
|
| 1760 |
+
"logits/rejected": 2.8990261554718018,
|
| 1761 |
+
"logps/chosen": -3.6797282695770264,
|
| 1762 |
+
"logps/rejected": -8.219030380249023,
|
| 1763 |
+
"loss": 0.2089,
|
| 1764 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1765 |
+
"rewards/chosen": -5.51959228515625,
|
| 1766 |
+
"rewards/margins": 6.808953285217285,
|
| 1767 |
+
"rewards/rejected": -12.328545570373535,
|
| 1768 |
+
"step": 970
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.8703374777975134,
|
| 1772 |
+
"grad_norm": 2.3526487350463867,
|
| 1773 |
+
"learning_rate": 1.3417599122003464e-06,
|
| 1774 |
+
"logits/chosen": 3.1998744010925293,
|
| 1775 |
+
"logits/rejected": 2.904794692993164,
|
| 1776 |
+
"logps/chosen": -4.254674911499023,
|
| 1777 |
+
"logps/rejected": -8.549466133117676,
|
| 1778 |
+
"loss": 0.1936,
|
| 1779 |
+
"rewards/accuracies": 0.9375,
|
| 1780 |
+
"rewards/chosen": -6.382012367248535,
|
| 1781 |
+
"rewards/margins": 6.4421868324279785,
|
| 1782 |
+
"rewards/rejected": -12.824198722839355,
|
| 1783 |
+
"step": 980
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.8792184724689165,
|
| 1787 |
+
"grad_norm": 2.372730016708374,
|
| 1788 |
+
"learning_rate": 1.2956158147457116e-06,
|
| 1789 |
+
"logits/chosen": 2.3626601696014404,
|
| 1790 |
+
"logits/rejected": 2.0850844383239746,
|
| 1791 |
+
"logps/chosen": -4.239551544189453,
|
| 1792 |
+
"logps/rejected": -8.695666313171387,
|
| 1793 |
+
"loss": 0.1902,
|
| 1794 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1795 |
+
"rewards/chosen": -6.359326362609863,
|
| 1796 |
+
"rewards/margins": 6.684172630310059,
|
| 1797 |
+
"rewards/rejected": -13.043498039245605,
|
| 1798 |
+
"step": 990
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.8880994671403197,
|
| 1802 |
+
"grad_norm": 3.703148365020752,
|
| 1803 |
+
"learning_rate": 1.2500000000000007e-06,
|
| 1804 |
+
"logits/chosen": 2.5446887016296387,
|
| 1805 |
+
"logits/rejected": 2.2514405250549316,
|
| 1806 |
+
"logps/chosen": -4.357522010803223,
|
| 1807 |
+
"logps/rejected": -8.732219696044922,
|
| 1808 |
+
"loss": 0.1807,
|
| 1809 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1810 |
+
"rewards/chosen": -6.53628396987915,
|
| 1811 |
+
"rewards/margins": 6.562047004699707,
|
| 1812 |
+
"rewards/rejected": -13.0983304977417,
|
| 1813 |
+
"step": 1000
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.8880994671403197,
|
| 1817 |
+
"eval_logits/chosen": 2.6361231803894043,
|
| 1818 |
+
"eval_logits/rejected": 2.4129557609558105,
|
| 1819 |
+
"eval_logps/chosen": -4.334481239318848,
|
| 1820 |
+
"eval_logps/rejected": -8.665815353393555,
|
| 1821 |
+
"eval_loss": 0.22103577852249146,
|
| 1822 |
+
"eval_rewards/accuracies": 0.9560439586639404,
|
| 1823 |
+
"eval_rewards/chosen": -6.501720905303955,
|
| 1824 |
+
"eval_rewards/margins": 6.497002601623535,
|
| 1825 |
+
"eval_rewards/rejected": -12.998723030090332,
|
| 1826 |
+
"eval_runtime": 25.2662,
|
| 1827 |
+
"eval_samples_per_second": 28.813,
|
| 1828 |
+
"eval_steps_per_second": 3.602,
|
| 1829 |
+
"step": 1000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.8969804618117229,
|
| 1833 |
+
"grad_norm": 3.239999532699585,
|
| 1834 |
+
"learning_rate": 1.204932476567175e-06,
|
| 1835 |
+
"logits/chosen": 2.8413162231445312,
|
| 1836 |
+
"logits/rejected": 2.5842719078063965,
|
| 1837 |
+
"logps/chosen": -3.9835076332092285,
|
| 1838 |
+
"logps/rejected": -8.900358200073242,
|
| 1839 |
+
"loss": 0.2276,
|
| 1840 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1841 |
+
"rewards/chosen": -5.975261688232422,
|
| 1842 |
+
"rewards/margins": 7.375277042388916,
|
| 1843 |
+
"rewards/rejected": -13.35053825378418,
|
| 1844 |
+
"step": 1010
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.9058614564831261,
|
| 1848 |
+
"grad_norm": 3.8074100017547607,
|
| 1849 |
+
"learning_rate": 1.160433012552508e-06,
|
| 1850 |
+
"logits/chosen": 1.8096859455108643,
|
| 1851 |
+
"logits/rejected": 1.5240850448608398,
|
| 1852 |
+
"logps/chosen": -4.373281002044678,
|
| 1853 |
+
"logps/rejected": -8.784773826599121,
|
| 1854 |
+
"loss": 0.1907,
|
| 1855 |
+
"rewards/accuracies": 0.9375,
|
| 1856 |
+
"rewards/chosen": -6.559922218322754,
|
| 1857 |
+
"rewards/margins": 6.6172380447387695,
|
| 1858 |
+
"rewards/rejected": -13.177160263061523,
|
| 1859 |
+
"step": 1020
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.9147424511545293,
|
| 1863 |
+
"grad_norm": 3.3684775829315186,
|
| 1864 |
+
"learning_rate": 1.11652112689164e-06,
|
| 1865 |
+
"logits/chosen": 2.5850119590759277,
|
| 1866 |
+
"logits/rejected": 2.503263235092163,
|
| 1867 |
+
"logps/chosen": -4.1612935066223145,
|
| 1868 |
+
"logps/rejected": -9.02672004699707,
|
| 1869 |
+
"loss": 0.2483,
|
| 1870 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1871 |
+
"rewards/chosen": -6.241940498352051,
|
| 1872 |
+
"rewards/margins": 7.298140525817871,
|
| 1873 |
+
"rewards/rejected": -13.540081977844238,
|
| 1874 |
+
"step": 1030
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.9236234458259325,
|
| 1878 |
+
"grad_norm": 4.323958873748779,
|
| 1879 |
+
"learning_rate": 1.073216080788921e-06,
|
| 1880 |
+
"logits/chosen": 3.1654298305511475,
|
| 1881 |
+
"logits/rejected": 2.52232027053833,
|
| 1882 |
+
"logps/chosen": -4.379122734069824,
|
| 1883 |
+
"logps/rejected": -8.86447525024414,
|
| 1884 |
+
"loss": 0.1995,
|
| 1885 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1886 |
+
"rewards/chosen": -6.5686845779418945,
|
| 1887 |
+
"rewards/margins": 6.728028774261475,
|
| 1888 |
+
"rewards/rejected": -13.296712875366211,
|
| 1889 |
+
"step": 1040
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.9325044404973357,
|
| 1893 |
+
"grad_norm": 2.923391342163086,
|
| 1894 |
+
"learning_rate": 1.0305368692688175e-06,
|
| 1895 |
+
"logits/chosen": 3.426027774810791,
|
| 1896 |
+
"logits/rejected": 2.8201966285705566,
|
| 1897 |
+
"logps/chosen": -4.2101850509643555,
|
| 1898 |
+
"logps/rejected": -8.720430374145508,
|
| 1899 |
+
"loss": 0.2275,
|
| 1900 |
+
"rewards/accuracies": 0.9375,
|
| 1901 |
+
"rewards/chosen": -6.315278053283691,
|
| 1902 |
+
"rewards/margins": 6.765368461608887,
|
| 1903 |
+
"rewards/rejected": -13.080645561218262,
|
| 1904 |
+
"step": 1050
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.9325044404973357,
|
| 1908 |
+
"eval_logits/chosen": 2.7177271842956543,
|
| 1909 |
+
"eval_logits/rejected": 2.45237135887146,
|
| 1910 |
+
"eval_logps/chosen": -4.310559272766113,
|
| 1911 |
+
"eval_logps/rejected": -8.67324447631836,
|
| 1912 |
+
"eval_loss": 0.2168363630771637,
|
| 1913 |
+
"eval_rewards/accuracies": 0.9670329689979553,
|
| 1914 |
+
"eval_rewards/chosen": -6.465838432312012,
|
| 1915 |
+
"eval_rewards/margins": 6.544029235839844,
|
| 1916 |
+
"eval_rewards/rejected": -13.009867668151855,
|
| 1917 |
+
"eval_runtime": 25.2426,
|
| 1918 |
+
"eval_samples_per_second": 28.84,
|
| 1919 |
+
"eval_steps_per_second": 3.605,
|
| 1920 |
+
"step": 1050
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.9413854351687388,
|
| 1924 |
+
"grad_norm": 2.5395636558532715,
|
| 1925 |
+
"learning_rate": 9.88502212844063e-07,
|
| 1926 |
+
"logits/chosen": 2.634228467941284,
|
| 1927 |
+
"logits/rejected": 2.447514772415161,
|
| 1928 |
+
"logps/chosen": -4.163097381591797,
|
| 1929 |
+
"logps/rejected": -9.294473648071289,
|
| 1930 |
+
"loss": 0.214,
|
| 1931 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1932 |
+
"rewards/chosen": -6.244646072387695,
|
| 1933 |
+
"rewards/margins": 7.697064399719238,
|
| 1934 |
+
"rewards/rejected": -13.9417085647583,
|
| 1935 |
+
"step": 1060
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.9502664298401421,
|
| 1939 |
+
"grad_norm": 2.4016919136047363,
|
| 1940 |
+
"learning_rate": 9.471305493042243e-07,
|
| 1941 |
+
"logits/chosen": 3.249488115310669,
|
| 1942 |
+
"logits/rejected": 3.138061285018921,
|
| 1943 |
+
"logps/chosen": -4.317265510559082,
|
| 1944 |
+
"logps/rejected": -9.449271202087402,
|
| 1945 |
+
"loss": 0.2135,
|
| 1946 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1947 |
+
"rewards/chosen": -6.475898742675781,
|
| 1948 |
+
"rewards/margins": 7.6980085372924805,
|
| 1949 |
+
"rewards/rejected": -14.173907279968262,
|
| 1950 |
+
"step": 1070
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.9591474245115453,
|
| 1954 |
+
"grad_norm": 7.170376300811768,
|
| 1955 |
+
"learning_rate": 9.064400256282757e-07,
|
| 1956 |
+
"logits/chosen": 2.5836658477783203,
|
| 1957 |
+
"logits/rejected": 2.584625482559204,
|
| 1958 |
+
"logps/chosen": -3.731645107269287,
|
| 1959 |
+
"logps/rejected": -8.913705825805664,
|
| 1960 |
+
"loss": 0.2002,
|
| 1961 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1962 |
+
"rewards/chosen": -5.597468376159668,
|
| 1963 |
+
"rewards/margins": 7.7730913162231445,
|
| 1964 |
+
"rewards/rejected": -13.370559692382812,
|
| 1965 |
+
"step": 1080
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.9680284191829485,
|
| 1969 |
+
"grad_norm": 4.712115287780762,
|
| 1970 |
+
"learning_rate": 8.664484900247363e-07,
|
| 1971 |
+
"logits/chosen": 1.7256263494491577,
|
| 1972 |
+
"logits/rejected": 1.679078459739685,
|
| 1973 |
+
"logps/chosen": -4.077349662780762,
|
| 1974 |
+
"logps/rejected": -9.539873123168945,
|
| 1975 |
+
"loss": 0.2498,
|
| 1976 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1977 |
+
"rewards/chosen": -6.116024494171143,
|
| 1978 |
+
"rewards/margins": 8.193784713745117,
|
| 1979 |
+
"rewards/rejected": -14.309808731079102,
|
| 1980 |
+
"step": 1090
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 0.9769094138543517,
|
| 1984 |
+
"grad_norm": 4.168649673461914,
|
| 1985 |
+
"learning_rate": 8.271734841028553e-07,
|
| 1986 |
+
"logits/chosen": 3.224684953689575,
|
| 1987 |
+
"logits/rejected": 2.831458568572998,
|
| 1988 |
+
"logps/chosen": -4.505261421203613,
|
| 1989 |
+
"logps/rejected": -9.360811233520508,
|
| 1990 |
+
"loss": 0.1854,
|
| 1991 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1992 |
+
"rewards/chosen": -6.757891654968262,
|
| 1993 |
+
"rewards/margins": 7.283324241638184,
|
| 1994 |
+
"rewards/rejected": -14.041216850280762,
|
| 1995 |
+
"step": 1100
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.9769094138543517,
|
| 1999 |
+
"eval_logits/chosen": 2.7460875511169434,
|
| 2000 |
+
"eval_logits/rejected": 2.4983255863189697,
|
| 2001 |
+
"eval_logps/chosen": -4.352819919586182,
|
| 2002 |
+
"eval_logps/rejected": -8.75997257232666,
|
| 2003 |
+
"eval_loss": 0.2113826721906662,
|
| 2004 |
+
"eval_rewards/accuracies": 0.9780219793319702,
|
| 2005 |
+
"eval_rewards/chosen": -6.529229640960693,
|
| 2006 |
+
"eval_rewards/margins": 6.610728740692139,
|
| 2007 |
+
"eval_rewards/rejected": -13.139957427978516,
|
| 2008 |
+
"eval_runtime": 25.2523,
|
| 2009 |
+
"eval_samples_per_second": 28.829,
|
| 2010 |
+
"eval_steps_per_second": 3.604,
|
| 2011 |
+
"step": 1100
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.9857904085257548,
|
| 2015 |
+
"grad_norm": 2.9944748878479004,
|
| 2016 |
+
"learning_rate": 7.886322351782782e-07,
|
| 2017 |
+
"logits/chosen": 2.92256236076355,
|
| 2018 |
+
"logits/rejected": 2.4897289276123047,
|
| 2019 |
+
"logps/chosen": -4.083096027374268,
|
| 2020 |
+
"logps/rejected": -8.914036750793457,
|
| 2021 |
+
"loss": 0.1986,
|
| 2022 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2023 |
+
"rewards/chosen": -6.124643325805664,
|
| 2024 |
+
"rewards/margins": 7.2464118003845215,
|
| 2025 |
+
"rewards/rejected": -13.371055603027344,
|
| 2026 |
+
"step": 1110
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.9946714031971581,
|
| 2030 |
+
"grad_norm": 3.546858310699463,
|
| 2031 |
+
"learning_rate": 7.508416487165862e-07,
|
| 2032 |
+
"logits/chosen": 2.6310508251190186,
|
| 2033 |
+
"logits/rejected": 2.64538836479187,
|
| 2034 |
+
"logps/chosen": -4.345538139343262,
|
| 2035 |
+
"logps/rejected": -9.072227478027344,
|
| 2036 |
+
"loss": 0.1663,
|
| 2037 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2038 |
+
"rewards/chosen": -6.518306732177734,
|
| 2039 |
+
"rewards/margins": 7.090034484863281,
|
| 2040 |
+
"rewards/rejected": -13.6083402633667,
|
| 2041 |
+
"step": 1120
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 1.0035523978685612,
|
| 2045 |
+
"grad_norm": 2.745328903198242,
|
| 2046 |
+
"learning_rate": 7.138183009179922e-07,
|
| 2047 |
+
"logits/chosen": 2.1050074100494385,
|
| 2048 |
+
"logits/rejected": 2.0917954444885254,
|
| 2049 |
+
"logps/chosen": -3.694542646408081,
|
| 2050 |
+
"logps/rejected": -8.503544807434082,
|
| 2051 |
+
"loss": 0.1985,
|
| 2052 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2053 |
+
"rewards/chosen": -5.541813850402832,
|
| 2054 |
+
"rewards/margins": 7.213502407073975,
|
| 2055 |
+
"rewards/rejected": -12.755315780639648,
|
| 2056 |
+
"step": 1130
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 1.0124333925399644,
|
| 2060 |
+
"grad_norm": 1.995431661605835,
|
| 2061 |
+
"learning_rate": 6.775784314464717e-07,
|
| 2062 |
+
"logits/chosen": 2.7968056201934814,
|
| 2063 |
+
"logits/rejected": 2.488126754760742,
|
| 2064 |
+
"logps/chosen": -4.206582546234131,
|
| 2065 |
+
"logps/rejected": -8.700386047363281,
|
| 2066 |
+
"loss": 0.2141,
|
| 2067 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2068 |
+
"rewards/chosen": -6.309874534606934,
|
| 2069 |
+
"rewards/margins": 6.740703582763672,
|
| 2070 |
+
"rewards/rejected": -13.050578117370605,
|
| 2071 |
+
"step": 1140
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 1.0213143872113677,
|
| 2075 |
+
"grad_norm": 3.110943078994751,
|
| 2076 |
+
"learning_rate": 6.421379363065142e-07,
|
| 2077 |
+
"logits/chosen": 2.5061817169189453,
|
| 2078 |
+
"logits/rejected": 2.342125415802002,
|
| 2079 |
+
"logps/chosen": -4.2504048347473145,
|
| 2080 |
+
"logps/rejected": -8.307449340820312,
|
| 2081 |
+
"loss": 0.2069,
|
| 2082 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2083 |
+
"rewards/chosen": -6.375607013702393,
|
| 2084 |
+
"rewards/margins": 6.085566520690918,
|
| 2085 |
+
"rewards/rejected": -12.461172103881836,
|
| 2086 |
+
"step": 1150
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 1.0213143872113677,
|
| 2090 |
+
"eval_logits/chosen": 2.728573799133301,
|
| 2091 |
+
"eval_logits/rejected": 2.5072758197784424,
|
| 2092 |
+
"eval_logps/chosen": -4.397165298461914,
|
| 2093 |
+
"eval_logps/rejected": -8.84023380279541,
|
| 2094 |
+
"eval_loss": 0.20922726392745972,
|
| 2095 |
+
"eval_rewards/accuracies": 0.9780219793319702,
|
| 2096 |
+
"eval_rewards/chosen": -6.595747947692871,
|
| 2097 |
+
"eval_rewards/margins": 6.664604663848877,
|
| 2098 |
+
"eval_rewards/rejected": -13.26035213470459,
|
| 2099 |
+
"eval_runtime": 25.264,
|
| 2100 |
+
"eval_samples_per_second": 28.816,
|
| 2101 |
+
"eval_steps_per_second": 3.602,
|
| 2102 |
+
"step": 1150
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.030195381882771,
|
| 2106 |
+
"grad_norm": 3.401503086090088,
|
| 2107 |
+
"learning_rate": 6.075123608706093e-07,
|
| 2108 |
+
"logits/chosen": 2.933258533477783,
|
| 2109 |
+
"logits/rejected": 2.5766117572784424,
|
| 2110 |
+
"logps/chosen": -4.0520782470703125,
|
| 2111 |
+
"logps/rejected": -9.317255973815918,
|
| 2112 |
+
"loss": 0.2031,
|
| 2113 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2114 |
+
"rewards/chosen": -6.078117847442627,
|
| 2115 |
+
"rewards/margins": 7.897767066955566,
|
| 2116 |
+
"rewards/rejected": -13.975885391235352,
|
| 2117 |
+
"step": 1160
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.0390763765541742,
|
| 2121 |
+
"grad_norm": 2.563494920730591,
|
| 2122 |
+
"learning_rate": 5.737168930605272e-07,
|
| 2123 |
+
"logits/chosen": 2.8566126823425293,
|
| 2124 |
+
"logits/rejected": 2.501013994216919,
|
| 2125 |
+
"logps/chosen": -4.048529148101807,
|
| 2126 |
+
"logps/rejected": -8.981905937194824,
|
| 2127 |
+
"loss": 0.2078,
|
| 2128 |
+
"rewards/accuracies": 0.9375,
|
| 2129 |
+
"rewards/chosen": -6.072793483734131,
|
| 2130 |
+
"rewards/margins": 7.400063991546631,
|
| 2131 |
+
"rewards/rejected": -13.472857475280762,
|
| 2132 |
+
"step": 1170
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 1.0479573712255772,
|
| 2136 |
+
"grad_norm": 8.221929550170898,
|
| 2137 |
+
"learning_rate": 5.407663566854008e-07,
|
| 2138 |
+
"logits/chosen": 1.8076190948486328,
|
| 2139 |
+
"logits/rejected": 1.5266883373260498,
|
| 2140 |
+
"logps/chosen": -3.697718858718872,
|
| 2141 |
+
"logps/rejected": -8.30528450012207,
|
| 2142 |
+
"loss": 0.1869,
|
| 2143 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2144 |
+
"rewards/chosen": -5.546578407287598,
|
| 2145 |
+
"rewards/margins": 6.911346435546875,
|
| 2146 |
+
"rewards/rejected": -12.457925796508789,
|
| 2147 |
+
"step": 1180
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 1.0568383658969804,
|
| 2151 |
+
"grad_norm": 6.9954400062561035,
|
| 2152 |
+
"learning_rate": 5.086752049395094e-07,
|
| 2153 |
+
"logits/chosen": 2.3529181480407715,
|
| 2154 |
+
"logits/rejected": 2.250333786010742,
|
| 2155 |
+
"logps/chosen": -4.018799781799316,
|
| 2156 |
+
"logps/rejected": -8.714046478271484,
|
| 2157 |
+
"loss": 0.2026,
|
| 2158 |
+
"rewards/accuracies": 0.9375,
|
| 2159 |
+
"rewards/chosen": -6.028199195861816,
|
| 2160 |
+
"rewards/margins": 7.042870998382568,
|
| 2161 |
+
"rewards/rejected": -13.071070671081543,
|
| 2162 |
+
"step": 1190
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 1.0657193605683837,
|
| 2166 |
+
"grad_norm": 3.6167593002319336,
|
| 2167 |
+
"learning_rate": 4.774575140626317e-07,
|
| 2168 |
+
"logits/chosen": 2.7552218437194824,
|
| 2169 |
+
"logits/rejected": 2.428560733795166,
|
| 2170 |
+
"logps/chosen": -4.324695587158203,
|
| 2171 |
+
"logps/rejected": -9.764104843139648,
|
| 2172 |
+
"loss": 0.2233,
|
| 2173 |
+
"rewards/accuracies": 0.9375,
|
| 2174 |
+
"rewards/chosen": -6.487043857574463,
|
| 2175 |
+
"rewards/margins": 8.159114837646484,
|
| 2176 |
+
"rewards/rejected": -14.646158218383789,
|
| 2177 |
+
"step": 1200
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 1.0657193605683837,
|
| 2181 |
+
"eval_logits/chosen": 2.7005860805511475,
|
| 2182 |
+
"eval_logits/rejected": 2.4944183826446533,
|
| 2183 |
+
"eval_logps/chosen": -4.283793926239014,
|
| 2184 |
+
"eval_logps/rejected": -8.801348686218262,
|
| 2185 |
+
"eval_loss": 0.20932108163833618,
|
| 2186 |
+
"eval_rewards/accuracies": 0.9560439586639404,
|
| 2187 |
+
"eval_rewards/chosen": -6.425690650939941,
|
| 2188 |
+
"eval_rewards/margins": 6.776332855224609,
|
| 2189 |
+
"eval_rewards/rejected": -13.202025413513184,
|
| 2190 |
+
"eval_runtime": 25.2698,
|
| 2191 |
+
"eval_samples_per_second": 28.809,
|
| 2192 |
+
"eval_steps_per_second": 3.601,
|
| 2193 |
+
"step": 1200
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.074600355239787,
|
| 2197 |
+
"grad_norm": 3.2438127994537354,
|
| 2198 |
+
"learning_rate": 4.4712697716573994e-07,
|
| 2199 |
+
"logits/chosen": 1.9486421346664429,
|
| 2200 |
+
"logits/rejected": 1.7296040058135986,
|
| 2201 |
+
"logps/chosen": -3.9212806224823,
|
| 2202 |
+
"logps/rejected": -8.304320335388184,
|
| 2203 |
+
"loss": 0.1761,
|
| 2204 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2205 |
+
"rewards/chosen": -5.881921291351318,
|
| 2206 |
+
"rewards/margins": 6.574559688568115,
|
| 2207 |
+
"rewards/rejected": -12.45648193359375,
|
| 2208 |
+
"step": 1210
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.0834813499111902,
|
| 2212 |
+
"grad_norm": 2.984710454940796,
|
| 2213 |
+
"learning_rate": 4.1769689822475147e-07,
|
| 2214 |
+
"logits/chosen": 3.56829833984375,
|
| 2215 |
+
"logits/rejected": 2.9676673412323,
|
| 2216 |
+
"logps/chosen": -4.453071117401123,
|
| 2217 |
+
"logps/rejected": -8.549017906188965,
|
| 2218 |
+
"loss": 0.1915,
|
| 2219 |
+
"rewards/accuracies": 0.9375,
|
| 2220 |
+
"rewards/chosen": -6.6796064376831055,
|
| 2221 |
+
"rewards/margins": 6.1439208984375,
|
| 2222 |
+
"rewards/rejected": -12.823527336120605,
|
| 2223 |
+
"step": 1220
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 1.0923623445825932,
|
| 2227 |
+
"grad_norm": 1.0289857387542725,
|
| 2228 |
+
"learning_rate": 3.891801862449629e-07,
|
| 2229 |
+
"logits/chosen": 2.529848098754883,
|
| 2230 |
+
"logits/rejected": 2.3582332134246826,
|
| 2231 |
+
"logps/chosen": -3.964491367340088,
|
| 2232 |
+
"logps/rejected": -9.1004638671875,
|
| 2233 |
+
"loss": 0.1743,
|
| 2234 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2235 |
+
"rewards/chosen": -5.946737289428711,
|
| 2236 |
+
"rewards/margins": 7.703960418701172,
|
| 2237 |
+
"rewards/rejected": -13.650697708129883,
|
| 2238 |
+
"step": 1230
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 1.1012433392539964,
|
| 2242 |
+
"grad_norm": 3.8722472190856934,
|
| 2243 |
+
"learning_rate": 3.615893495987335e-07,
|
| 2244 |
+
"logits/chosen": 2.980180263519287,
|
| 2245 |
+
"logits/rejected": 2.823392152786255,
|
| 2246 |
+
"logps/chosen": -4.560516357421875,
|
| 2247 |
+
"logps/rejected": -9.179193496704102,
|
| 2248 |
+
"loss": 0.195,
|
| 2249 |
+
"rewards/accuracies": 0.925000011920929,
|
| 2250 |
+
"rewards/chosen": -6.8407745361328125,
|
| 2251 |
+
"rewards/margins": 6.928016662597656,
|
| 2252 |
+
"rewards/rejected": -13.768791198730469,
|
| 2253 |
+
"step": 1240
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 1.1101243339253997,
|
| 2257 |
+
"grad_norm": 3.892000198364258,
|
| 2258 |
+
"learning_rate": 3.3493649053890325e-07,
|
| 2259 |
+
"logits/chosen": 2.8612444400787354,
|
| 2260 |
+
"logits/rejected": 2.5222373008728027,
|
| 2261 |
+
"logps/chosen": -4.044814586639404,
|
| 2262 |
+
"logps/rejected": -8.645524978637695,
|
| 2263 |
+
"loss": 0.1571,
|
| 2264 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2265 |
+
"rewards/chosen": -6.067221641540527,
|
| 2266 |
+
"rewards/margins": 6.901064872741699,
|
| 2267 |
+
"rewards/rejected": -12.968287467956543,
|
| 2268 |
+
"step": 1250
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.1101243339253997,
|
| 2272 |
+
"eval_logits/chosen": 2.714733839035034,
|
| 2273 |
+
"eval_logits/rejected": 2.5066633224487305,
|
| 2274 |
+
"eval_logps/chosen": -4.303951263427734,
|
| 2275 |
+
"eval_logps/rejected": -8.823600769042969,
|
| 2276 |
+
"eval_loss": 0.2080618292093277,
|
| 2277 |
+
"eval_rewards/accuracies": 0.9560439586639404,
|
| 2278 |
+
"eval_rewards/chosen": -6.45592737197876,
|
| 2279 |
+
"eval_rewards/margins": 6.77947473526001,
|
| 2280 |
+
"eval_rewards/rejected": -13.235400199890137,
|
| 2281 |
+
"eval_runtime": 25.2645,
|
| 2282 |
+
"eval_samples_per_second": 28.815,
|
| 2283 |
+
"eval_steps_per_second": 3.602,
|
| 2284 |
+
"step": 1250
|
| 2285 |
+
}
|
| 2286 |
+
],
|
| 2287 |
+
"logging_steps": 10,
|
| 2288 |
+
"max_steps": 1500,
|
| 2289 |
+
"num_input_tokens_seen": 0,
|
| 2290 |
+
"num_train_epochs": 2,
|
| 2291 |
+
"save_steps": 50,
|
| 2292 |
+
"stateful_callbacks": {
|
| 2293 |
+
"TrainerControl": {
|
| 2294 |
+
"args": {
|
| 2295 |
+
"should_epoch_stop": false,
|
| 2296 |
+
"should_evaluate": false,
|
| 2297 |
+
"should_log": false,
|
| 2298 |
+
"should_save": true,
|
| 2299 |
+
"should_training_stop": false
|
| 2300 |
+
},
|
| 2301 |
+
"attributes": {}
|
| 2302 |
+
}
|
| 2303 |
+
},
|
| 2304 |
+
"total_flos": 2.9419979120889037e+18,
|
| 2305 |
+
"train_batch_size": 1,
|
| 2306 |
+
"trial_name": null,
|
| 2307 |
+
"trial_params": null
|
| 2308 |
+
}
|
checkpoint-1250/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1a9f338a2c311ad932497119af4f36e393c3e914b2390e2dd2e31575d923108
|
| 3 |
+
size 7224
|
checkpoint-1250/zero_to_fp32.py
ADDED
|
@@ -0,0 +1,674 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python
|
| 2 |
+
|
| 3 |
+
# Copyright (c) Microsoft Corporation.
|
| 4 |
+
# SPDX-License-Identifier: Apache-2.0
|
| 5 |
+
|
| 6 |
+
# DeepSpeed Team
|
| 7 |
+
|
| 8 |
+
# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
|
| 9 |
+
# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
|
| 10 |
+
# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
|
| 11 |
+
# application.
|
| 12 |
+
#
|
| 13 |
+
# example:
|
| 14 |
+
# python zero_to_fp32.py . output_dir/
|
| 15 |
+
# or
|
| 16 |
+
# python zero_to_fp32.py . output_dir/ --safe_serialization
|
| 17 |
+
|
| 18 |
+
import argparse
|
| 19 |
+
import torch
|
| 20 |
+
import glob
|
| 21 |
+
import math
|
| 22 |
+
import os
|
| 23 |
+
import re
|
| 24 |
+
import json
|
| 25 |
+
from tqdm import tqdm
|
| 26 |
+
from collections import OrderedDict
|
| 27 |
+
from dataclasses import dataclass
|
| 28 |
+
|
| 29 |
+
# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
|
| 30 |
+
# DeepSpeed data structures it has to be available in the current python environment.
|
| 31 |
+
from deepspeed.utils import logger
|
| 32 |
+
from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
|
| 33 |
+
FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
|
| 34 |
+
FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
@dataclass
|
| 38 |
+
class zero_model_state:
|
| 39 |
+
buffers: dict()
|
| 40 |
+
param_shapes: dict()
|
| 41 |
+
shared_params: list
|
| 42 |
+
ds_version: int
|
| 43 |
+
frozen_param_shapes: dict()
|
| 44 |
+
frozen_param_fragments: dict()
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
debug = 0
|
| 48 |
+
|
| 49 |
+
# load to cpu
|
| 50 |
+
device = torch.device('cpu')
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
def atoi(text):
|
| 54 |
+
return int(text) if text.isdigit() else text
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
def natural_keys(text):
|
| 58 |
+
'''
|
| 59 |
+
alist.sort(key=natural_keys) sorts in human order
|
| 60 |
+
http://nedbatchelder.com/blog/200712/human_sorting.html
|
| 61 |
+
(See Toothy's implementation in the comments)
|
| 62 |
+
'''
|
| 63 |
+
return [atoi(c) for c in re.split(r'(\d+)', text)]
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
def get_model_state_file(checkpoint_dir, zero_stage):
|
| 67 |
+
if not os.path.isdir(checkpoint_dir):
|
| 68 |
+
raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
|
| 69 |
+
|
| 70 |
+
# there should be only one file
|
| 71 |
+
if zero_stage <= 2:
|
| 72 |
+
file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
|
| 73 |
+
elif zero_stage == 3:
|
| 74 |
+
file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
|
| 75 |
+
|
| 76 |
+
if not os.path.exists(file):
|
| 77 |
+
raise FileNotFoundError(f"can't find model states file at '{file}'")
|
| 78 |
+
|
| 79 |
+
return file
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
def get_checkpoint_files(checkpoint_dir, glob_pattern):
|
| 83 |
+
# XXX: need to test that this simple glob rule works for multi-node setup too
|
| 84 |
+
ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
|
| 85 |
+
|
| 86 |
+
if len(ckpt_files) == 0:
|
| 87 |
+
raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
|
| 88 |
+
|
| 89 |
+
return ckpt_files
|
| 90 |
+
|
| 91 |
+
|
| 92 |
+
def get_optim_files(checkpoint_dir):
|
| 93 |
+
return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
|
| 94 |
+
|
| 95 |
+
|
| 96 |
+
def get_model_state_files(checkpoint_dir):
|
| 97 |
+
return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
def parse_model_states(files):
|
| 101 |
+
zero_model_states = []
|
| 102 |
+
for file in files:
|
| 103 |
+
state_dict = torch.load(file, map_location=device)
|
| 104 |
+
|
| 105 |
+
if BUFFER_NAMES not in state_dict:
|
| 106 |
+
raise ValueError(f"{file} is not a model state checkpoint")
|
| 107 |
+
buffer_names = state_dict[BUFFER_NAMES]
|
| 108 |
+
if debug:
|
| 109 |
+
print("Found buffers:", buffer_names)
|
| 110 |
+
|
| 111 |
+
# recover just the buffers while restoring them to fp32 if they were saved in fp16
|
| 112 |
+
buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
|
| 113 |
+
param_shapes = state_dict[PARAM_SHAPES]
|
| 114 |
+
|
| 115 |
+
# collect parameters that are included in param_shapes
|
| 116 |
+
param_names = []
|
| 117 |
+
for s in param_shapes:
|
| 118 |
+
for name in s.keys():
|
| 119 |
+
param_names.append(name)
|
| 120 |
+
|
| 121 |
+
# update with frozen parameters
|
| 122 |
+
frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
|
| 123 |
+
if frozen_param_shapes is not None:
|
| 124 |
+
if debug:
|
| 125 |
+
print(f"Found frozen_param_shapes: {frozen_param_shapes}")
|
| 126 |
+
param_names += list(frozen_param_shapes.keys())
|
| 127 |
+
|
| 128 |
+
# handle shared params
|
| 129 |
+
shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
|
| 130 |
+
|
| 131 |
+
ds_version = state_dict.get(DS_VERSION, None)
|
| 132 |
+
|
| 133 |
+
frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
|
| 134 |
+
|
| 135 |
+
z_model_state = zero_model_state(buffers=buffers,
|
| 136 |
+
param_shapes=param_shapes,
|
| 137 |
+
shared_params=shared_params,
|
| 138 |
+
ds_version=ds_version,
|
| 139 |
+
frozen_param_shapes=frozen_param_shapes,
|
| 140 |
+
frozen_param_fragments=frozen_param_fragments)
|
| 141 |
+
zero_model_states.append(z_model_state)
|
| 142 |
+
|
| 143 |
+
return zero_model_states
|
| 144 |
+
|
| 145 |
+
|
| 146 |
+
def parse_optim_states(files, ds_checkpoint_dir):
|
| 147 |
+
total_files = len(files)
|
| 148 |
+
state_dicts = []
|
| 149 |
+
for f in files:
|
| 150 |
+
state_dict = torch.load(f, map_location=device)
|
| 151 |
+
# immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
|
| 152 |
+
# and also handle the case where it was already removed by another helper script
|
| 153 |
+
state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
|
| 154 |
+
state_dicts.append(state_dict)
|
| 155 |
+
|
| 156 |
+
if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
|
| 157 |
+
raise ValueError(f"{files[0]} is not a zero checkpoint")
|
| 158 |
+
zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
|
| 159 |
+
world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
|
| 160 |
+
|
| 161 |
+
# For ZeRO-2 each param group can have different partition_count as data parallelism for expert
|
| 162 |
+
# parameters can be different from data parallelism for non-expert parameters. So we can just
|
| 163 |
+
# use the max of the partition_count to get the dp world_size.
|
| 164 |
+
|
| 165 |
+
if type(world_size) is list:
|
| 166 |
+
world_size = max(world_size)
|
| 167 |
+
|
| 168 |
+
if world_size != total_files:
|
| 169 |
+
raise ValueError(
|
| 170 |
+
f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
|
| 171 |
+
"Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
|
| 172 |
+
)
|
| 173 |
+
|
| 174 |
+
# the groups are named differently in each stage
|
| 175 |
+
if zero_stage <= 2:
|
| 176 |
+
fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
|
| 177 |
+
elif zero_stage == 3:
|
| 178 |
+
fp32_groups_key = FP32_FLAT_GROUPS
|
| 179 |
+
else:
|
| 180 |
+
raise ValueError(f"unknown zero stage {zero_stage}")
|
| 181 |
+
|
| 182 |
+
if zero_stage <= 2:
|
| 183 |
+
fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
|
| 184 |
+
elif zero_stage == 3:
|
| 185 |
+
# if there is more than one param group, there will be multiple flattened tensors - one
|
| 186 |
+
# flattened tensor per group - for simplicity merge them into a single tensor
|
| 187 |
+
#
|
| 188 |
+
# XXX: could make the script more memory efficient for when there are multiple groups - it
|
| 189 |
+
# will require matching the sub-lists of param_shapes for each param group flattened tensor
|
| 190 |
+
|
| 191 |
+
fp32_flat_groups = [
|
| 192 |
+
torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
|
| 193 |
+
]
|
| 194 |
+
|
| 195 |
+
return zero_stage, world_size, fp32_flat_groups
|
| 196 |
+
|
| 197 |
+
|
| 198 |
+
def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
|
| 199 |
+
"""
|
| 200 |
+
Returns fp32 state_dict reconstructed from ds checkpoint
|
| 201 |
+
|
| 202 |
+
Args:
|
| 203 |
+
- ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
|
| 204 |
+
|
| 205 |
+
"""
|
| 206 |
+
print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
|
| 207 |
+
|
| 208 |
+
optim_files = get_optim_files(ds_checkpoint_dir)
|
| 209 |
+
zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
|
| 210 |
+
print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
|
| 211 |
+
|
| 212 |
+
model_files = get_model_state_files(ds_checkpoint_dir)
|
| 213 |
+
|
| 214 |
+
zero_model_states = parse_model_states(model_files)
|
| 215 |
+
print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
|
| 216 |
+
|
| 217 |
+
if zero_stage <= 2:
|
| 218 |
+
return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 219 |
+
exclude_frozen_parameters)
|
| 220 |
+
elif zero_stage == 3:
|
| 221 |
+
return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 222 |
+
exclude_frozen_parameters)
|
| 223 |
+
|
| 224 |
+
|
| 225 |
+
def _zero2_merge_frozen_params(state_dict, zero_model_states):
|
| 226 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 227 |
+
return
|
| 228 |
+
|
| 229 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 230 |
+
frozen_param_fragments = zero_model_states[0].frozen_param_fragments
|
| 231 |
+
|
| 232 |
+
if debug:
|
| 233 |
+
num_elem = sum(s.numel() for s in frozen_param_shapes.values())
|
| 234 |
+
print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 235 |
+
|
| 236 |
+
wanted_params = len(frozen_param_shapes)
|
| 237 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 238 |
+
avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
|
| 239 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 240 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 241 |
+
|
| 242 |
+
total_params = 0
|
| 243 |
+
total_numel = 0
|
| 244 |
+
for name, shape in frozen_param_shapes.items():
|
| 245 |
+
total_params += 1
|
| 246 |
+
unpartitioned_numel = shape.numel()
|
| 247 |
+
total_numel += unpartitioned_numel
|
| 248 |
+
|
| 249 |
+
state_dict[name] = frozen_param_fragments[name]
|
| 250 |
+
|
| 251 |
+
if debug:
|
| 252 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 253 |
+
|
| 254 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 255 |
+
|
| 256 |
+
|
| 257 |
+
def _has_callable(obj, fn):
|
| 258 |
+
attr = getattr(obj, fn, None)
|
| 259 |
+
return callable(attr)
|
| 260 |
+
|
| 261 |
+
|
| 262 |
+
def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 263 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 264 |
+
|
| 265 |
+
# Reconstruction protocol:
|
| 266 |
+
#
|
| 267 |
+
# XXX: document this
|
| 268 |
+
|
| 269 |
+
if debug:
|
| 270 |
+
for i in range(world_size):
|
| 271 |
+
for j in range(len(fp32_flat_groups[0])):
|
| 272 |
+
print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
|
| 273 |
+
|
| 274 |
+
# XXX: memory usage doubles here (zero2)
|
| 275 |
+
num_param_groups = len(fp32_flat_groups[0])
|
| 276 |
+
merged_single_partition_of_fp32_groups = []
|
| 277 |
+
for i in range(num_param_groups):
|
| 278 |
+
merged_partitions = [sd[i] for sd in fp32_flat_groups]
|
| 279 |
+
full_single_fp32_vector = torch.cat(merged_partitions, 0)
|
| 280 |
+
merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
|
| 281 |
+
avail_numel = sum(
|
| 282 |
+
[full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
|
| 283 |
+
|
| 284 |
+
if debug:
|
| 285 |
+
wanted_params = sum([len(shapes) for shapes in param_shapes])
|
| 286 |
+
wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
|
| 287 |
+
# not asserting if there is a mismatch due to possible padding
|
| 288 |
+
print(f"Have {avail_numel} numels to process.")
|
| 289 |
+
print(f"Need {wanted_numel} numels in {wanted_params} params.")
|
| 290 |
+
|
| 291 |
+
# params
|
| 292 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 293 |
+
# out-of-core computing solution
|
| 294 |
+
total_numel = 0
|
| 295 |
+
total_params = 0
|
| 296 |
+
for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
|
| 297 |
+
offset = 0
|
| 298 |
+
avail_numel = full_single_fp32_vector.numel()
|
| 299 |
+
for name, shape in shapes.items():
|
| 300 |
+
|
| 301 |
+
unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
|
| 302 |
+
total_numel += unpartitioned_numel
|
| 303 |
+
total_params += 1
|
| 304 |
+
|
| 305 |
+
if debug:
|
| 306 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 307 |
+
state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
|
| 308 |
+
offset += unpartitioned_numel
|
| 309 |
+
|
| 310 |
+
# Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
|
| 311 |
+
# avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
|
| 312 |
+
# paddings performed in the code it's almost impossible to predict the exact numbers w/o the
|
| 313 |
+
# live optimizer object, so we are checking that the numbers are within the right range
|
| 314 |
+
align_to = 2 * world_size
|
| 315 |
+
|
| 316 |
+
def zero2_align(x):
|
| 317 |
+
return align_to * math.ceil(x / align_to)
|
| 318 |
+
|
| 319 |
+
if debug:
|
| 320 |
+
print(f"original offset={offset}, avail_numel={avail_numel}")
|
| 321 |
+
|
| 322 |
+
offset = zero2_align(offset)
|
| 323 |
+
avail_numel = zero2_align(avail_numel)
|
| 324 |
+
|
| 325 |
+
if debug:
|
| 326 |
+
print(f"aligned offset={offset}, avail_numel={avail_numel}")
|
| 327 |
+
|
| 328 |
+
# Sanity check
|
| 329 |
+
if offset != avail_numel:
|
| 330 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 331 |
+
|
| 332 |
+
print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
|
| 333 |
+
|
| 334 |
+
|
| 335 |
+
def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 336 |
+
exclude_frozen_parameters):
|
| 337 |
+
state_dict = OrderedDict()
|
| 338 |
+
|
| 339 |
+
# buffers
|
| 340 |
+
buffers = zero_model_states[0].buffers
|
| 341 |
+
state_dict.update(buffers)
|
| 342 |
+
if debug:
|
| 343 |
+
print(f"added {len(buffers)} buffers")
|
| 344 |
+
|
| 345 |
+
if not exclude_frozen_parameters:
|
| 346 |
+
_zero2_merge_frozen_params(state_dict, zero_model_states)
|
| 347 |
+
|
| 348 |
+
_zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 349 |
+
|
| 350 |
+
# recover shared parameters
|
| 351 |
+
for pair in zero_model_states[0].shared_params:
|
| 352 |
+
if pair[1] in state_dict:
|
| 353 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 354 |
+
|
| 355 |
+
return state_dict
|
| 356 |
+
|
| 357 |
+
|
| 358 |
+
def zero3_partitioned_param_info(unpartitioned_numel, world_size):
|
| 359 |
+
remainder = unpartitioned_numel % world_size
|
| 360 |
+
padding_numel = (world_size - remainder) if remainder else 0
|
| 361 |
+
partitioned_numel = math.ceil(unpartitioned_numel / world_size)
|
| 362 |
+
return partitioned_numel, padding_numel
|
| 363 |
+
|
| 364 |
+
|
| 365 |
+
def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
|
| 366 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 367 |
+
return
|
| 368 |
+
|
| 369 |
+
if debug:
|
| 370 |
+
for i in range(world_size):
|
| 371 |
+
num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
|
| 372 |
+
print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 373 |
+
|
| 374 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 375 |
+
wanted_params = len(frozen_param_shapes)
|
| 376 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 377 |
+
avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
|
| 378 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 379 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 380 |
+
|
| 381 |
+
total_params = 0
|
| 382 |
+
total_numel = 0
|
| 383 |
+
for name, shape in zero_model_states[0].frozen_param_shapes.items():
|
| 384 |
+
total_params += 1
|
| 385 |
+
unpartitioned_numel = shape.numel()
|
| 386 |
+
total_numel += unpartitioned_numel
|
| 387 |
+
|
| 388 |
+
param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
|
| 389 |
+
state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 390 |
+
|
| 391 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 392 |
+
|
| 393 |
+
if debug:
|
| 394 |
+
print(
|
| 395 |
+
f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 396 |
+
)
|
| 397 |
+
|
| 398 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 399 |
+
|
| 400 |
+
|
| 401 |
+
def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 402 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 403 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 404 |
+
# Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
|
| 405 |
+
# param, re-consolidating each param, while dealing with padding if any
|
| 406 |
+
|
| 407 |
+
# merge list of dicts, preserving order
|
| 408 |
+
param_shapes = {k: v for d in param_shapes for k, v in d.items()}
|
| 409 |
+
|
| 410 |
+
if debug:
|
| 411 |
+
for i in range(world_size):
|
| 412 |
+
print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
|
| 413 |
+
|
| 414 |
+
wanted_params = len(param_shapes)
|
| 415 |
+
wanted_numel = sum(shape.numel() for shape in param_shapes.values())
|
| 416 |
+
# not asserting if there is a mismatch due to possible padding
|
| 417 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 418 |
+
print(f"Trainable params: Have {avail_numel} numels to process.")
|
| 419 |
+
print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
|
| 420 |
+
|
| 421 |
+
# params
|
| 422 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 423 |
+
# out-of-core computing solution
|
| 424 |
+
offset = 0
|
| 425 |
+
total_numel = 0
|
| 426 |
+
total_params = 0
|
| 427 |
+
for name, shape in tqdm(param_shapes.items(), desc='Gathering Sharded Weights'):
|
| 428 |
+
unpartitioned_numel = shape.numel()
|
| 429 |
+
total_numel += unpartitioned_numel
|
| 430 |
+
total_params += 1
|
| 431 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 432 |
+
|
| 433 |
+
if debug:
|
| 434 |
+
print(
|
| 435 |
+
f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 436 |
+
)
|
| 437 |
+
|
| 438 |
+
# XXX: memory usage doubles here
|
| 439 |
+
state_dict[name] = torch.cat(
|
| 440 |
+
tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
|
| 441 |
+
0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 442 |
+
offset += partitioned_numel
|
| 443 |
+
|
| 444 |
+
offset *= world_size
|
| 445 |
+
|
| 446 |
+
# Sanity check
|
| 447 |
+
if offset != avail_numel:
|
| 448 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 449 |
+
|
| 450 |
+
print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
|
| 451 |
+
|
| 452 |
+
|
| 453 |
+
def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 454 |
+
exclude_frozen_parameters):
|
| 455 |
+
state_dict = OrderedDict()
|
| 456 |
+
|
| 457 |
+
# buffers
|
| 458 |
+
buffers = zero_model_states[0].buffers
|
| 459 |
+
state_dict.update(buffers)
|
| 460 |
+
if debug:
|
| 461 |
+
print(f"added {len(buffers)} buffers")
|
| 462 |
+
|
| 463 |
+
if not exclude_frozen_parameters:
|
| 464 |
+
_zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
|
| 465 |
+
|
| 466 |
+
_zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 467 |
+
|
| 468 |
+
# recover shared parameters
|
| 469 |
+
for pair in zero_model_states[0].shared_params:
|
| 470 |
+
if pair[1] in state_dict:
|
| 471 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 472 |
+
|
| 473 |
+
return state_dict
|
| 474 |
+
|
| 475 |
+
|
| 476 |
+
def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
|
| 477 |
+
"""
|
| 478 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
|
| 479 |
+
``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
|
| 480 |
+
via a model hub.
|
| 481 |
+
|
| 482 |
+
Args:
|
| 483 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder
|
| 484 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
|
| 485 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 486 |
+
|
| 487 |
+
Returns:
|
| 488 |
+
- pytorch ``state_dict``
|
| 489 |
+
|
| 490 |
+
Note: this approach may not work if your application doesn't have sufficient free CPU memory and
|
| 491 |
+
you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
|
| 492 |
+
the checkpoint.
|
| 493 |
+
|
| 494 |
+
A typical usage might be ::
|
| 495 |
+
|
| 496 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 497 |
+
# do the training and checkpoint saving
|
| 498 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
|
| 499 |
+
model = model.cpu() # move to cpu
|
| 500 |
+
model.load_state_dict(state_dict)
|
| 501 |
+
# submit to model hub or save the model to share with others
|
| 502 |
+
|
| 503 |
+
In this example the ``model`` will no longer be usable in the deepspeed context of the same
|
| 504 |
+
application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 505 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 506 |
+
|
| 507 |
+
If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
|
| 508 |
+
|
| 509 |
+
"""
|
| 510 |
+
if tag is None:
|
| 511 |
+
latest_path = os.path.join(checkpoint_dir, 'latest')
|
| 512 |
+
if os.path.isfile(latest_path):
|
| 513 |
+
with open(latest_path, 'r') as fd:
|
| 514 |
+
tag = fd.read().strip()
|
| 515 |
+
else:
|
| 516 |
+
raise ValueError(f"Unable to find 'latest' file at {latest_path}")
|
| 517 |
+
|
| 518 |
+
ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
|
| 519 |
+
|
| 520 |
+
if not os.path.isdir(ds_checkpoint_dir):
|
| 521 |
+
raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
|
| 522 |
+
|
| 523 |
+
return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
|
| 524 |
+
|
| 525 |
+
|
| 526 |
+
def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
|
| 527 |
+
output_dir,
|
| 528 |
+
max_shard_size="5GB",
|
| 529 |
+
safe_serialization=False,
|
| 530 |
+
tag=None,
|
| 531 |
+
exclude_frozen_parameters=False):
|
| 532 |
+
"""
|
| 533 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
|
| 534 |
+
loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
|
| 535 |
+
|
| 536 |
+
Args:
|
| 537 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 538 |
+
- ``output_dir``: directory to the pytorch fp32 state_dict output files
|
| 539 |
+
- ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
|
| 540 |
+
- ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
|
| 541 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 542 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 543 |
+
"""
|
| 544 |
+
# Dependency pre-check
|
| 545 |
+
if safe_serialization:
|
| 546 |
+
try:
|
| 547 |
+
from safetensors.torch import save_file
|
| 548 |
+
except ImportError:
|
| 549 |
+
print('If you want to use `safe_serialization`, please `pip install safetensors`')
|
| 550 |
+
raise
|
| 551 |
+
if max_shard_size is not None:
|
| 552 |
+
try:
|
| 553 |
+
from huggingface_hub import split_torch_state_dict_into_shards
|
| 554 |
+
except ImportError:
|
| 555 |
+
print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
|
| 556 |
+
raise
|
| 557 |
+
|
| 558 |
+
# Convert zero checkpoint to state_dict
|
| 559 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
|
| 560 |
+
|
| 561 |
+
# Shard the model if it is too big.
|
| 562 |
+
weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
|
| 563 |
+
if max_shard_size is not None:
|
| 564 |
+
filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
|
| 565 |
+
state_dict_split = split_torch_state_dict_into_shards(state_dict,
|
| 566 |
+
filename_pattern=filename_pattern,
|
| 567 |
+
max_shard_size=max_shard_size)
|
| 568 |
+
else:
|
| 569 |
+
from collections import namedtuple
|
| 570 |
+
StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
|
| 571 |
+
state_dict_split = StateDictSplit(is_sharded=False,
|
| 572 |
+
filename_to_tensors={weights_name: list(state_dict.keys())})
|
| 573 |
+
|
| 574 |
+
# Save the model
|
| 575 |
+
filename_to_tensors = state_dict_split.filename_to_tensors.items()
|
| 576 |
+
for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
|
| 577 |
+
shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
|
| 578 |
+
output_path = os.path.join(output_dir, shard_file)
|
| 579 |
+
if safe_serialization:
|
| 580 |
+
save_file(shard, output_path, metadata={"format": "pt"})
|
| 581 |
+
else:
|
| 582 |
+
torch.save(shard, output_path)
|
| 583 |
+
|
| 584 |
+
# Save index if sharded
|
| 585 |
+
if state_dict_split.is_sharded:
|
| 586 |
+
index = {
|
| 587 |
+
"metadata": state_dict_split.metadata,
|
| 588 |
+
"weight_map": state_dict_split.tensor_to_filename,
|
| 589 |
+
}
|
| 590 |
+
save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
|
| 591 |
+
save_index_file = os.path.join(output_dir, save_index_file)
|
| 592 |
+
with open(save_index_file, "w", encoding="utf-8") as f:
|
| 593 |
+
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
|
| 594 |
+
f.write(content)
|
| 595 |
+
|
| 596 |
+
|
| 597 |
+
def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
|
| 598 |
+
"""
|
| 599 |
+
1. Put the provided model to cpu
|
| 600 |
+
2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
|
| 601 |
+
3. Load it into the provided model
|
| 602 |
+
|
| 603 |
+
Args:
|
| 604 |
+
- ``model``: the model object to update
|
| 605 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 606 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 607 |
+
|
| 608 |
+
Returns:
|
| 609 |
+
- ``model`: modified model
|
| 610 |
+
|
| 611 |
+
Make sure you have plenty of CPU memory available before you call this function. If you don't
|
| 612 |
+
have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
|
| 613 |
+
conveniently placed for you in the checkpoint folder.
|
| 614 |
+
|
| 615 |
+
A typical usage might be ::
|
| 616 |
+
|
| 617 |
+
from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
|
| 618 |
+
model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
|
| 619 |
+
# submit to model hub or save the model to share with others
|
| 620 |
+
|
| 621 |
+
Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
|
| 622 |
+
of the same application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 623 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 624 |
+
|
| 625 |
+
"""
|
| 626 |
+
logger.info(f"Extracting fp32 weights")
|
| 627 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
|
| 628 |
+
|
| 629 |
+
logger.info(f"Overwriting model with fp32 weights")
|
| 630 |
+
model = model.cpu()
|
| 631 |
+
model.load_state_dict(state_dict, strict=False)
|
| 632 |
+
|
| 633 |
+
return model
|
| 634 |
+
|
| 635 |
+
|
| 636 |
+
if __name__ == "__main__":
|
| 637 |
+
parser = argparse.ArgumentParser()
|
| 638 |
+
parser.add_argument("checkpoint_dir",
|
| 639 |
+
type=str,
|
| 640 |
+
help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
|
| 641 |
+
parser.add_argument("output_dir",
|
| 642 |
+
type=str,
|
| 643 |
+
help="directory to the pytorch fp32 state_dict output files"
|
| 644 |
+
"(e.g. path/checkpoint-12-output/)")
|
| 645 |
+
parser.add_argument(
|
| 646 |
+
"--max_shard_size",
|
| 647 |
+
type=str,
|
| 648 |
+
default="5GB",
|
| 649 |
+
help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
|
| 650 |
+
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
|
| 651 |
+
"We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
|
| 652 |
+
"without CPU OOM issues.")
|
| 653 |
+
parser.add_argument(
|
| 654 |
+
"--safe_serialization",
|
| 655 |
+
default=False,
|
| 656 |
+
action='store_true',
|
| 657 |
+
help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
|
| 658 |
+
parser.add_argument("-t",
|
| 659 |
+
"--tag",
|
| 660 |
+
type=str,
|
| 661 |
+
default=None,
|
| 662 |
+
help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
|
| 663 |
+
parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
|
| 664 |
+
parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
|
| 665 |
+
args = parser.parse_args()
|
| 666 |
+
|
| 667 |
+
debug = args.debug
|
| 668 |
+
|
| 669 |
+
convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
|
| 670 |
+
args.output_dir,
|
| 671 |
+
max_shard_size=args.max_shard_size,
|
| 672 |
+
safe_serialization=args.safe_serialization,
|
| 673 |
+
tag=args.tag,
|
| 674 |
+
exclude_frozen_parameters=args.exclude_frozen_parameters)
|