Upload folder using huggingface_hub
Browse files- README.md +10 -0
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- config.json +45 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +1074 -0
README.md
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
### Framework versions
|
| 7 |
+
|
| 8 |
+
- PEFT 0.4.0
|
| 9 |
+
|
| 10 |
+
- PEFT 0.4.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "lmsys/vicuna-13b-v1.5",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 256,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 128,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"k_proj",
|
| 19 |
+
"up_proj",
|
| 20 |
+
"v_proj",
|
| 21 |
+
"gate_proj",
|
| 22 |
+
"down_proj",
|
| 23 |
+
"o_proj"
|
| 24 |
+
],
|
| 25 |
+
"task_type": "CAUSAL_LM"
|
| 26 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b202492563a54e11e38c2fe8a50b787f1e6d2371650728f0f77043411c8f8d2
|
| 3 |
+
size 1001584778
|
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "lmsys/vicuna-13b-v1.5",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"freeze_mm_mlp_adapter": false,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 5120,
|
| 13 |
+
"image_aspect_ratio": "pad",
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 13824,
|
| 16 |
+
"max_length": 4096,
|
| 17 |
+
"max_position_embeddings": 4096,
|
| 18 |
+
"mm_hidden_size": 1024,
|
| 19 |
+
"mm_patch_merge_type": "flat",
|
| 20 |
+
"mm_projector_lr": 2e-05,
|
| 21 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 22 |
+
"mm_use_im_patch_token": false,
|
| 23 |
+
"mm_use_im_start_end": false,
|
| 24 |
+
"mm_vision_select_feature": "patch",
|
| 25 |
+
"mm_vision_select_layer": -2,
|
| 26 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
| 27 |
+
"model_type": "llava_llama",
|
| 28 |
+
"num_attention_heads": 40,
|
| 29 |
+
"num_hidden_layers": 40,
|
| 30 |
+
"num_key_value_heads": 40,
|
| 31 |
+
"pad_token_id": 0,
|
| 32 |
+
"pretraining_tp": 1,
|
| 33 |
+
"rms_norm_eps": 1e-05,
|
| 34 |
+
"rope_scaling": null,
|
| 35 |
+
"rope_theta": 10000.0,
|
| 36 |
+
"tie_word_embeddings": false,
|
| 37 |
+
"tokenizer_model_max_length": 2048,
|
| 38 |
+
"tokenizer_padding_side": "right",
|
| 39 |
+
"torch_dtype": "float16",
|
| 40 |
+
"transformers_version": "4.37.2",
|
| 41 |
+
"tune_mm_mlp_adapter": false,
|
| 42 |
+
"use_cache": true,
|
| 43 |
+
"use_mm_proj": true,
|
| 44 |
+
"vocab_size": 32000
|
| 45 |
+
}
|
non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3bbe7bfb9feaf3ec5da7190da5abbefb3cdcc80e81b56ef76c83e4a049bf40e
|
| 3 |
+
size 62937264
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1074 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 174,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 14 |
+
"loss": 3.0336,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 6.666666666666667e-05,
|
| 20 |
+
"loss": 3.1471,
|
| 21 |
+
"step": 2
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.02,
|
| 25 |
+
"learning_rate": 0.0001,
|
| 26 |
+
"loss": 2.5604,
|
| 27 |
+
"step": 3
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 0.00013333333333333334,
|
| 32 |
+
"loss": 2.0848,
|
| 33 |
+
"step": 4
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 0.0001666666666666667,
|
| 38 |
+
"loss": 1.8695,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.03,
|
| 43 |
+
"learning_rate": 0.0002,
|
| 44 |
+
"loss": 1.6627,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04,
|
| 49 |
+
"learning_rate": 0.00019998251609127464,
|
| 50 |
+
"loss": 1.6042,
|
| 51 |
+
"step": 7
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05,
|
| 55 |
+
"learning_rate": 0.00019993007047883988,
|
| 56 |
+
"loss": 1.6334,
|
| 57 |
+
"step": 8
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.05,
|
| 61 |
+
"learning_rate": 0.00019984268150178167,
|
| 62 |
+
"loss": 1.5498,
|
| 63 |
+
"step": 9
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.06,
|
| 67 |
+
"learning_rate": 0.00019972037971811802,
|
| 68 |
+
"loss": 1.4238,
|
| 69 |
+
"step": 10
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.06,
|
| 73 |
+
"learning_rate": 0.00019956320789411337,
|
| 74 |
+
"loss": 1.5104,
|
| 75 |
+
"step": 11
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.07,
|
| 79 |
+
"learning_rate": 0.00019937122098932428,
|
| 80 |
+
"loss": 1.399,
|
| 81 |
+
"step": 12
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.07,
|
| 85 |
+
"learning_rate": 0.00019914448613738106,
|
| 86 |
+
"loss": 1.4097,
|
| 87 |
+
"step": 13
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.08,
|
| 91 |
+
"learning_rate": 0.00019888308262251285,
|
| 92 |
+
"loss": 1.2895,
|
| 93 |
+
"step": 14
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09,
|
| 97 |
+
"learning_rate": 0.0001985871018518236,
|
| 98 |
+
"loss": 1.2686,
|
| 99 |
+
"step": 15
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.09,
|
| 103 |
+
"learning_rate": 0.00019825664732332884,
|
| 104 |
+
"loss": 1.3434,
|
| 105 |
+
"step": 16
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.1,
|
| 109 |
+
"learning_rate": 0.00019789183458976484,
|
| 110 |
+
"loss": 1.3547,
|
| 111 |
+
"step": 17
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.1,
|
| 115 |
+
"learning_rate": 0.00019749279121818235,
|
| 116 |
+
"loss": 1.2858,
|
| 117 |
+
"step": 18
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.11,
|
| 121 |
+
"learning_rate": 0.0001970596567453391,
|
| 122 |
+
"loss": 1.2182,
|
| 123 |
+
"step": 19
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.11,
|
| 127 |
+
"learning_rate": 0.00019659258262890683,
|
| 128 |
+
"loss": 1.2566,
|
| 129 |
+
"step": 20
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.12,
|
| 133 |
+
"learning_rate": 0.00019609173219450998,
|
| 134 |
+
"loss": 1.1689,
|
| 135 |
+
"step": 21
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.13,
|
| 139 |
+
"learning_rate": 0.0001955572805786141,
|
| 140 |
+
"loss": 1.2791,
|
| 141 |
+
"step": 22
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.13,
|
| 145 |
+
"learning_rate": 0.0001949894146672846,
|
| 146 |
+
"loss": 1.2545,
|
| 147 |
+
"step": 23
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.14,
|
| 151 |
+
"learning_rate": 0.00019438833303083678,
|
| 152 |
+
"loss": 1.2585,
|
| 153 |
+
"step": 24
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.14,
|
| 157 |
+
"learning_rate": 0.00019375424585439994,
|
| 158 |
+
"loss": 1.155,
|
| 159 |
+
"step": 25
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.15,
|
| 163 |
+
"learning_rate": 0.00019308737486442045,
|
| 164 |
+
"loss": 1.2322,
|
| 165 |
+
"step": 26
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.16,
|
| 169 |
+
"learning_rate": 0.0001923879532511287,
|
| 170 |
+
"loss": 1.0583,
|
| 171 |
+
"step": 27
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.16,
|
| 175 |
+
"learning_rate": 0.00019165622558699763,
|
| 176 |
+
"loss": 1.0604,
|
| 177 |
+
"step": 28
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17,
|
| 181 |
+
"learning_rate": 0.0001908924477412211,
|
| 182 |
+
"loss": 1.1765,
|
| 183 |
+
"step": 29
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17,
|
| 187 |
+
"learning_rate": 0.0001900968867902419,
|
| 188 |
+
"loss": 1.0841,
|
| 189 |
+
"step": 30
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.18,
|
| 193 |
+
"learning_rate": 0.00018926982092436117,
|
| 194 |
+
"loss": 1.2448,
|
| 195 |
+
"step": 31
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.18,
|
| 199 |
+
"learning_rate": 0.00018841153935046098,
|
| 200 |
+
"loss": 1.2845,
|
| 201 |
+
"step": 32
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.19,
|
| 205 |
+
"learning_rate": 0.00018752234219087538,
|
| 206 |
+
"loss": 1.0631,
|
| 207 |
+
"step": 33
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.2,
|
| 211 |
+
"learning_rate": 0.00018660254037844388,
|
| 212 |
+
"loss": 1.1415,
|
| 213 |
+
"step": 34
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2,
|
| 217 |
+
"learning_rate": 0.00018565245554778517,
|
| 218 |
+
"loss": 1.0635,
|
| 219 |
+
"step": 35
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.21,
|
| 223 |
+
"learning_rate": 0.00018467241992282843,
|
| 224 |
+
"loss": 1.118,
|
| 225 |
+
"step": 36
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.21,
|
| 229 |
+
"learning_rate": 0.00018366277620064197,
|
| 230 |
+
"loss": 1.125,
|
| 231 |
+
"step": 37
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.22,
|
| 235 |
+
"learning_rate": 0.0001826238774315995,
|
| 236 |
+
"loss": 1.0472,
|
| 237 |
+
"step": 38
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.22,
|
| 241 |
+
"learning_rate": 0.00018155608689592604,
|
| 242 |
+
"loss": 1.0681,
|
| 243 |
+
"step": 39
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.23,
|
| 247 |
+
"learning_rate": 0.00018045977797666684,
|
| 248 |
+
"loss": 1.0575,
|
| 249 |
+
"step": 40
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.24,
|
| 253 |
+
"learning_rate": 0.00017933533402912354,
|
| 254 |
+
"loss": 1.1236,
|
| 255 |
+
"step": 41
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.24,
|
| 259 |
+
"learning_rate": 0.000178183148246803,
|
| 260 |
+
"loss": 0.9568,
|
| 261 |
+
"step": 42
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.25,
|
| 265 |
+
"learning_rate": 0.0001770036235239263,
|
| 266 |
+
"loss": 0.9808,
|
| 267 |
+
"step": 43
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.25,
|
| 271 |
+
"learning_rate": 0.0001757971723145453,
|
| 272 |
+
"loss": 1.0124,
|
| 273 |
+
"step": 44
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.26,
|
| 277 |
+
"learning_rate": 0.00017456421648831655,
|
| 278 |
+
"loss": 1.0146,
|
| 279 |
+
"step": 45
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.26,
|
| 283 |
+
"learning_rate": 0.00017330518718298264,
|
| 284 |
+
"loss": 0.8494,
|
| 285 |
+
"step": 46
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.27,
|
| 289 |
+
"learning_rate": 0.00017202052465361268,
|
| 290 |
+
"loss": 0.9945,
|
| 291 |
+
"step": 47
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.28,
|
| 295 |
+
"learning_rate": 0.00017071067811865476,
|
| 296 |
+
"loss": 0.9945,
|
| 297 |
+
"step": 48
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.28,
|
| 301 |
+
"learning_rate": 0.00016937610560285418,
|
| 302 |
+
"loss": 0.9627,
|
| 303 |
+
"step": 49
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.29,
|
| 307 |
+
"learning_rate": 0.00016801727377709194,
|
| 308 |
+
"loss": 1.0944,
|
| 309 |
+
"step": 50
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.29,
|
| 313 |
+
"learning_rate": 0.0001666346577952004,
|
| 314 |
+
"loss": 0.9068,
|
| 315 |
+
"step": 51
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.3,
|
| 319 |
+
"learning_rate": 0.00016522874112781213,
|
| 320 |
+
"loss": 0.9744,
|
| 321 |
+
"step": 52
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.3,
|
| 325 |
+
"learning_rate": 0.0001638000153933009,
|
| 326 |
+
"loss": 0.9814,
|
| 327 |
+
"step": 53
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.31,
|
| 331 |
+
"learning_rate": 0.00016234898018587337,
|
| 332 |
+
"loss": 0.897,
|
| 333 |
+
"step": 54
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.32,
|
| 337 |
+
"learning_rate": 0.00016087614290087208,
|
| 338 |
+
"loss": 0.916,
|
| 339 |
+
"step": 55
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.32,
|
| 343 |
+
"learning_rate": 0.00015938201855735014,
|
| 344 |
+
"loss": 0.9574,
|
| 345 |
+
"step": 56
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.33,
|
| 349 |
+
"learning_rate": 0.0001578671296179806,
|
| 350 |
+
"loss": 0.9101,
|
| 351 |
+
"step": 57
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.33,
|
| 355 |
+
"learning_rate": 0.0001563320058063622,
|
| 356 |
+
"loss": 0.9506,
|
| 357 |
+
"step": 58
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.34,
|
| 361 |
+
"learning_rate": 0.00015477718392178716,
|
| 362 |
+
"loss": 0.8437,
|
| 363 |
+
"step": 59
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.34,
|
| 367 |
+
"learning_rate": 0.00015320320765153367,
|
| 368 |
+
"loss": 0.9731,
|
| 369 |
+
"step": 60
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.35,
|
| 373 |
+
"learning_rate": 0.00015161062738075067,
|
| 374 |
+
"loss": 0.8724,
|
| 375 |
+
"step": 61
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.36,
|
| 379 |
+
"learning_rate": 0.00015000000000000001,
|
| 380 |
+
"loss": 1.0002,
|
| 381 |
+
"step": 62
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.36,
|
| 385 |
+
"learning_rate": 0.000148371888710524,
|
| 386 |
+
"loss": 0.9179,
|
| 387 |
+
"step": 63
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.37,
|
| 391 |
+
"learning_rate": 0.0001467268628273062,
|
| 392 |
+
"loss": 0.9031,
|
| 393 |
+
"step": 64
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.37,
|
| 397 |
+
"learning_rate": 0.00014506549757999454,
|
| 398 |
+
"loss": 0.9408,
|
| 399 |
+
"step": 65
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.38,
|
| 403 |
+
"learning_rate": 0.00014338837391175582,
|
| 404 |
+
"loss": 0.8223,
|
| 405 |
+
"step": 66
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.39,
|
| 409 |
+
"learning_rate": 0.00014169607827613283,
|
| 410 |
+
"loss": 0.8489,
|
| 411 |
+
"step": 67
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.39,
|
| 415 |
+
"learning_rate": 0.00013998920243197407,
|
| 416 |
+
"loss": 0.857,
|
| 417 |
+
"step": 68
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.4,
|
| 421 |
+
"learning_rate": 0.000138268343236509,
|
| 422 |
+
"loss": 0.883,
|
| 423 |
+
"step": 69
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4,
|
| 427 |
+
"learning_rate": 0.00013653410243663952,
|
| 428 |
+
"loss": 0.8169,
|
| 429 |
+
"step": 70
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.41,
|
| 433 |
+
"learning_rate": 0.00013478708645852272,
|
| 434 |
+
"loss": 0.7553,
|
| 435 |
+
"step": 71
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.41,
|
| 439 |
+
"learning_rate": 0.00013302790619551674,
|
| 440 |
+
"loss": 0.8621,
|
| 441 |
+
"step": 72
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.42,
|
| 445 |
+
"learning_rate": 0.00013125717679456447,
|
| 446 |
+
"loss": 0.779,
|
| 447 |
+
"step": 73
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.43,
|
| 451 |
+
"learning_rate": 0.00012947551744109043,
|
| 452 |
+
"loss": 0.8823,
|
| 453 |
+
"step": 74
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.43,
|
| 457 |
+
"learning_rate": 0.00012768355114248494,
|
| 458 |
+
"loss": 0.8303,
|
| 459 |
+
"step": 75
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.44,
|
| 463 |
+
"learning_rate": 0.00012588190451025207,
|
| 464 |
+
"loss": 0.8004,
|
| 465 |
+
"step": 76
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.44,
|
| 469 |
+
"learning_rate": 0.00012407120754089732,
|
| 470 |
+
"loss": 0.7605,
|
| 471 |
+
"step": 77
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.45,
|
| 475 |
+
"learning_rate": 0.00012225209339563145,
|
| 476 |
+
"loss": 0.7432,
|
| 477 |
+
"step": 78
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.45,
|
| 481 |
+
"learning_rate": 0.00012042519817896804,
|
| 482 |
+
"loss": 0.8013,
|
| 483 |
+
"step": 79
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.46,
|
| 487 |
+
"learning_rate": 0.00011859116071629149,
|
| 488 |
+
"loss": 0.8088,
|
| 489 |
+
"step": 80
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.47,
|
| 493 |
+
"learning_rate": 0.00011675062233047364,
|
| 494 |
+
"loss": 0.834,
|
| 495 |
+
"step": 81
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.47,
|
| 499 |
+
"learning_rate": 0.00011490422661761744,
|
| 500 |
+
"loss": 0.7438,
|
| 501 |
+
"step": 82
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.48,
|
| 505 |
+
"learning_rate": 0.00011305261922200519,
|
| 506 |
+
"loss": 0.8594,
|
| 507 |
+
"step": 83
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.48,
|
| 511 |
+
"learning_rate": 0.00011119644761033078,
|
| 512 |
+
"loss": 0.7394,
|
| 513 |
+
"step": 84
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.49,
|
| 517 |
+
"learning_rate": 0.00010933636084529506,
|
| 518 |
+
"loss": 0.7285,
|
| 519 |
+
"step": 85
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.49,
|
| 523 |
+
"learning_rate": 0.00010747300935864243,
|
| 524 |
+
"loss": 0.8386,
|
| 525 |
+
"step": 86
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.5,
|
| 529 |
+
"learning_rate": 0.00010560704472371919,
|
| 530 |
+
"loss": 0.6397,
|
| 531 |
+
"step": 87
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.51,
|
| 535 |
+
"learning_rate": 0.0001037391194276326,
|
| 536 |
+
"loss": 0.6312,
|
| 537 |
+
"step": 88
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.51,
|
| 541 |
+
"learning_rate": 0.00010186988664309023,
|
| 542 |
+
"loss": 0.7131,
|
| 543 |
+
"step": 89
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.52,
|
| 547 |
+
"learning_rate": 0.0001,
|
| 548 |
+
"loss": 0.7044,
|
| 549 |
+
"step": 90
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.52,
|
| 553 |
+
"learning_rate": 9.813011335690981e-05,
|
| 554 |
+
"loss": 0.6069,
|
| 555 |
+
"step": 91
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.53,
|
| 559 |
+
"learning_rate": 9.626088057236745e-05,
|
| 560 |
+
"loss": 0.7156,
|
| 561 |
+
"step": 92
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.53,
|
| 565 |
+
"learning_rate": 9.439295527628081e-05,
|
| 566 |
+
"loss": 0.6827,
|
| 567 |
+
"step": 93
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.54,
|
| 571 |
+
"learning_rate": 9.252699064135758e-05,
|
| 572 |
+
"loss": 0.7909,
|
| 573 |
+
"step": 94
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.55,
|
| 577 |
+
"learning_rate": 9.066363915470495e-05,
|
| 578 |
+
"loss": 0.6702,
|
| 579 |
+
"step": 95
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.55,
|
| 583 |
+
"learning_rate": 8.880355238966923e-05,
|
| 584 |
+
"loss": 0.78,
|
| 585 |
+
"step": 96
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.56,
|
| 589 |
+
"learning_rate": 8.694738077799488e-05,
|
| 590 |
+
"loss": 0.6501,
|
| 591 |
+
"step": 97
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.56,
|
| 595 |
+
"learning_rate": 8.509577338238255e-05,
|
| 596 |
+
"loss": 0.7133,
|
| 597 |
+
"step": 98
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.57,
|
| 601 |
+
"learning_rate": 8.324937766952638e-05,
|
| 602 |
+
"loss": 0.6917,
|
| 603 |
+
"step": 99
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.57,
|
| 607 |
+
"learning_rate": 8.140883928370855e-05,
|
| 608 |
+
"loss": 0.6802,
|
| 609 |
+
"step": 100
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.58,
|
| 613 |
+
"learning_rate": 7.957480182103198e-05,
|
| 614 |
+
"loss": 0.6907,
|
| 615 |
+
"step": 101
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.59,
|
| 619 |
+
"learning_rate": 7.774790660436858e-05,
|
| 620 |
+
"loss": 0.5801,
|
| 621 |
+
"step": 102
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.59,
|
| 625 |
+
"learning_rate": 7.592879245910273e-05,
|
| 626 |
+
"loss": 0.6508,
|
| 627 |
+
"step": 103
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.6,
|
| 631 |
+
"learning_rate": 7.411809548974792e-05,
|
| 632 |
+
"loss": 0.7096,
|
| 633 |
+
"step": 104
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.6,
|
| 637 |
+
"learning_rate": 7.231644885751507e-05,
|
| 638 |
+
"loss": 0.6661,
|
| 639 |
+
"step": 105
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.61,
|
| 643 |
+
"learning_rate": 7.052448255890957e-05,
|
| 644 |
+
"loss": 0.6926,
|
| 645 |
+
"step": 106
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.61,
|
| 649 |
+
"learning_rate": 6.874282320543557e-05,
|
| 650 |
+
"loss": 0.6867,
|
| 651 |
+
"step": 107
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.62,
|
| 655 |
+
"learning_rate": 6.697209380448333e-05,
|
| 656 |
+
"loss": 0.6498,
|
| 657 |
+
"step": 108
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.63,
|
| 661 |
+
"learning_rate": 6.521291354147727e-05,
|
| 662 |
+
"loss": 0.6701,
|
| 663 |
+
"step": 109
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.63,
|
| 667 |
+
"learning_rate": 6.34658975633605e-05,
|
| 668 |
+
"loss": 0.6017,
|
| 669 |
+
"step": 110
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.64,
|
| 673 |
+
"learning_rate": 6.173165676349103e-05,
|
| 674 |
+
"loss": 0.6007,
|
| 675 |
+
"step": 111
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.64,
|
| 679 |
+
"learning_rate": 6.001079756802592e-05,
|
| 680 |
+
"loss": 0.6231,
|
| 681 |
+
"step": 112
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.65,
|
| 685 |
+
"learning_rate": 5.830392172386723e-05,
|
| 686 |
+
"loss": 0.6004,
|
| 687 |
+
"step": 113
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.66,
|
| 691 |
+
"learning_rate": 5.6611626088244194e-05,
|
| 692 |
+
"loss": 0.6116,
|
| 693 |
+
"step": 114
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.66,
|
| 697 |
+
"learning_rate": 5.493450242000546e-05,
|
| 698 |
+
"loss": 0.6014,
|
| 699 |
+
"step": 115
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.67,
|
| 703 |
+
"learning_rate": 5.32731371726938e-05,
|
| 704 |
+
"loss": 0.6195,
|
| 705 |
+
"step": 116
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.67,
|
| 709 |
+
"learning_rate": 5.162811128947602e-05,
|
| 710 |
+
"loss": 0.5959,
|
| 711 |
+
"step": 117
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.68,
|
| 715 |
+
"learning_rate": 5.000000000000002e-05,
|
| 716 |
+
"loss": 0.5329,
|
| 717 |
+
"step": 118
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.68,
|
| 721 |
+
"learning_rate": 4.8389372619249326e-05,
|
| 722 |
+
"loss": 0.5667,
|
| 723 |
+
"step": 119
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.69,
|
| 727 |
+
"learning_rate": 4.6796792348466356e-05,
|
| 728 |
+
"loss": 0.5541,
|
| 729 |
+
"step": 120
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.7,
|
| 733 |
+
"learning_rate": 4.522281607821288e-05,
|
| 734 |
+
"loss": 0.572,
|
| 735 |
+
"step": 121
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.7,
|
| 739 |
+
"learning_rate": 4.3667994193637796e-05,
|
| 740 |
+
"loss": 0.5608,
|
| 741 |
+
"step": 122
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.71,
|
| 745 |
+
"learning_rate": 4.213287038201943e-05,
|
| 746 |
+
"loss": 0.6407,
|
| 747 |
+
"step": 123
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.71,
|
| 751 |
+
"learning_rate": 4.0617981442649855e-05,
|
| 752 |
+
"loss": 0.6392,
|
| 753 |
+
"step": 124
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.72,
|
| 757 |
+
"learning_rate": 3.9123857099127936e-05,
|
| 758 |
+
"loss": 0.5541,
|
| 759 |
+
"step": 125
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.72,
|
| 763 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 764 |
+
"loss": 0.6149,
|
| 765 |
+
"step": 126
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.73,
|
| 769 |
+
"learning_rate": 3.6199984606699155e-05,
|
| 770 |
+
"loss": 0.5239,
|
| 771 |
+
"step": 127
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.74,
|
| 775 |
+
"learning_rate": 3.477125887218792e-05,
|
| 776 |
+
"loss": 0.6098,
|
| 777 |
+
"step": 128
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.74,
|
| 781 |
+
"learning_rate": 3.336534220479961e-05,
|
| 782 |
+
"loss": 0.5971,
|
| 783 |
+
"step": 129
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.75,
|
| 787 |
+
"learning_rate": 3.198272622290804e-05,
|
| 788 |
+
"loss": 0.5751,
|
| 789 |
+
"step": 130
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.75,
|
| 793 |
+
"learning_rate": 3.062389439714584e-05,
|
| 794 |
+
"loss": 0.5351,
|
| 795 |
+
"step": 131
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.76,
|
| 799 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 800 |
+
"loss": 0.5096,
|
| 801 |
+
"step": 132
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.76,
|
| 805 |
+
"learning_rate": 2.797947534638736e-05,
|
| 806 |
+
"loss": 0.5955,
|
| 807 |
+
"step": 133
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.77,
|
| 811 |
+
"learning_rate": 2.669481281701739e-05,
|
| 812 |
+
"loss": 0.5452,
|
| 813 |
+
"step": 134
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.78,
|
| 817 |
+
"learning_rate": 2.5435783511683443e-05,
|
| 818 |
+
"loss": 0.5363,
|
| 819 |
+
"step": 135
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.78,
|
| 823 |
+
"learning_rate": 2.420282768545469e-05,
|
| 824 |
+
"loss": 0.4915,
|
| 825 |
+
"step": 136
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.79,
|
| 829 |
+
"learning_rate": 2.2996376476073723e-05,
|
| 830 |
+
"loss": 0.5574,
|
| 831 |
+
"step": 137
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.79,
|
| 835 |
+
"learning_rate": 2.181685175319702e-05,
|
| 836 |
+
"loss": 0.5438,
|
| 837 |
+
"step": 138
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.8,
|
| 841 |
+
"learning_rate": 2.0664665970876496e-05,
|
| 842 |
+
"loss": 0.5464,
|
| 843 |
+
"step": 139
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.8,
|
| 847 |
+
"learning_rate": 1.9540222023333166e-05,
|
| 848 |
+
"loss": 0.5587,
|
| 849 |
+
"step": 140
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.81,
|
| 853 |
+
"learning_rate": 1.8443913104073983e-05,
|
| 854 |
+
"loss": 0.5535,
|
| 855 |
+
"step": 141
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.82,
|
| 859 |
+
"learning_rate": 1.7376122568400532e-05,
|
| 860 |
+
"loss": 0.5958,
|
| 861 |
+
"step": 142
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.82,
|
| 865 |
+
"learning_rate": 1.6337223799358026e-05,
|
| 866 |
+
"loss": 0.5997,
|
| 867 |
+
"step": 143
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.83,
|
| 871 |
+
"learning_rate": 1.5327580077171587e-05,
|
| 872 |
+
"loss": 0.5441,
|
| 873 |
+
"step": 144
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.83,
|
| 877 |
+
"learning_rate": 1.4347544452214868e-05,
|
| 878 |
+
"loss": 0.4945,
|
| 879 |
+
"step": 145
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.84,
|
| 883 |
+
"learning_rate": 1.339745962155613e-05,
|
| 884 |
+
"loss": 0.5564,
|
| 885 |
+
"step": 146
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.84,
|
| 889 |
+
"learning_rate": 1.2477657809124631e-05,
|
| 890 |
+
"loss": 0.4948,
|
| 891 |
+
"step": 147
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.85,
|
| 895 |
+
"learning_rate": 1.1588460649539035e-05,
|
| 896 |
+
"loss": 0.5315,
|
| 897 |
+
"step": 148
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.86,
|
| 901 |
+
"learning_rate": 1.0730179075638868e-05,
|
| 902 |
+
"loss": 0.5196,
|
| 903 |
+
"step": 149
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.86,
|
| 907 |
+
"learning_rate": 9.903113209758096e-06,
|
| 908 |
+
"loss": 0.5382,
|
| 909 |
+
"step": 150
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.87,
|
| 913 |
+
"learning_rate": 9.107552258778907e-06,
|
| 914 |
+
"loss": 0.55,
|
| 915 |
+
"step": 151
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.87,
|
| 919 |
+
"learning_rate": 8.343774413002381e-06,
|
| 920 |
+
"loss": 0.524,
|
| 921 |
+
"step": 152
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.88,
|
| 925 |
+
"learning_rate": 7.612046748871327e-06,
|
| 926 |
+
"loss": 0.5602,
|
| 927 |
+
"step": 153
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.89,
|
| 931 |
+
"learning_rate": 6.9126251355795864e-06,
|
| 932 |
+
"loss": 0.5391,
|
| 933 |
+
"step": 154
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.89,
|
| 937 |
+
"learning_rate": 6.245754145600091e-06,
|
| 938 |
+
"loss": 0.5259,
|
| 939 |
+
"step": 155
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.9,
|
| 943 |
+
"learning_rate": 5.611666969163243e-06,
|
| 944 |
+
"loss": 0.5483,
|
| 945 |
+
"step": 156
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.9,
|
| 949 |
+
"learning_rate": 5.0105853327154e-06,
|
| 950 |
+
"loss": 0.5533,
|
| 951 |
+
"step": 157
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.91,
|
| 955 |
+
"learning_rate": 4.442719421385922e-06,
|
| 956 |
+
"loss": 0.5403,
|
| 957 |
+
"step": 158
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.91,
|
| 961 |
+
"learning_rate": 3.908267805490051e-06,
|
| 962 |
+
"loss": 0.5192,
|
| 963 |
+
"step": 159
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.92,
|
| 967 |
+
"learning_rate": 3.40741737109318e-06,
|
| 968 |
+
"loss": 0.5134,
|
| 969 |
+
"step": 160
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.93,
|
| 973 |
+
"learning_rate": 2.9403432546609043e-06,
|
| 974 |
+
"loss": 0.4957,
|
| 975 |
+
"step": 161
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.93,
|
| 979 |
+
"learning_rate": 2.5072087818176382e-06,
|
| 980 |
+
"loss": 0.429,
|
| 981 |
+
"step": 162
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.94,
|
| 985 |
+
"learning_rate": 2.1081654102351635e-06,
|
| 986 |
+
"loss": 0.5649,
|
| 987 |
+
"step": 163
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.94,
|
| 991 |
+
"learning_rate": 1.7433526766711728e-06,
|
| 992 |
+
"loss": 0.5655,
|
| 993 |
+
"step": 164
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.95,
|
| 997 |
+
"learning_rate": 1.4128981481764115e-06,
|
| 998 |
+
"loss": 0.568,
|
| 999 |
+
"step": 165
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.95,
|
| 1003 |
+
"learning_rate": 1.1169173774871478e-06,
|
| 1004 |
+
"loss": 0.5024,
|
| 1005 |
+
"step": 166
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.96,
|
| 1009 |
+
"learning_rate": 8.555138626189618e-07,
|
| 1010 |
+
"loss": 0.5127,
|
| 1011 |
+
"step": 167
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.97,
|
| 1015 |
+
"learning_rate": 6.287790106757396e-07,
|
| 1016 |
+
"loss": 0.5509,
|
| 1017 |
+
"step": 168
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.97,
|
| 1021 |
+
"learning_rate": 4.3679210588661866e-07,
|
| 1022 |
+
"loss": 0.5531,
|
| 1023 |
+
"step": 169
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.98,
|
| 1027 |
+
"learning_rate": 2.7962028188198706e-07,
|
| 1028 |
+
"loss": 0.5437,
|
| 1029 |
+
"step": 170
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.98,
|
| 1033 |
+
"learning_rate": 1.5731849821833954e-07,
|
| 1034 |
+
"loss": 0.4631,
|
| 1035 |
+
"step": 171
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.99,
|
| 1039 |
+
"learning_rate": 6.992952116013918e-08,
|
| 1040 |
+
"loss": 0.5563,
|
| 1041 |
+
"step": 172
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.99,
|
| 1045 |
+
"learning_rate": 1.7483908725357545e-08,
|
| 1046 |
+
"loss": 0.5058,
|
| 1047 |
+
"step": 173
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.0,
|
| 1051 |
+
"learning_rate": 0.0,
|
| 1052 |
+
"loss": 0.5082,
|
| 1053 |
+
"step": 174
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.0,
|
| 1057 |
+
"step": 174,
|
| 1058 |
+
"total_flos": 4413720231936.0,
|
| 1059 |
+
"train_loss": 0.8584621692525929,
|
| 1060 |
+
"train_runtime": 3538.7692,
|
| 1061 |
+
"train_samples_per_second": 3.132,
|
| 1062 |
+
"train_steps_per_second": 0.049
|
| 1063 |
+
}
|
| 1064 |
+
],
|
| 1065 |
+
"logging_steps": 1.0,
|
| 1066 |
+
"max_steps": 174,
|
| 1067 |
+
"num_input_tokens_seen": 0,
|
| 1068 |
+
"num_train_epochs": 1,
|
| 1069 |
+
"save_steps": 50000,
|
| 1070 |
+
"total_flos": 4413720231936.0,
|
| 1071 |
+
"train_batch_size": 16,
|
| 1072 |
+
"trial_name": null,
|
| 1073 |
+
"trial_params": null
|
| 1074 |
+
}
|