Training in progress, step 20000, checkpoint
Browse files- checkpoint-20000/config.json +29 -0
- checkpoint-20000/generation_config.json +7 -0
- checkpoint-20000/model.safetensors +3 -0
- checkpoint-20000/optimizer.pt +3 -0
- checkpoint-20000/rng_state.pth +3 -0
- checkpoint-20000/scheduler.pt +3 -0
- checkpoint-20000/special_tokens_map.json +24 -0
- checkpoint-20000/tokenizer.json +0 -0
- checkpoint-20000/tokenizer.model +3 -0
- checkpoint-20000/tokenizer_config.json +45 -0
- checkpoint-20000/trainer_state.json +493 -0
- checkpoint-20000/training_args.bin +3 -0
checkpoint-20000/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "JackFram/llama-68m",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"model_type": "llama",
|
| 16 |
+
"num_attention_heads": 12,
|
| 17 |
+
"num_hidden_layers": 2,
|
| 18 |
+
"num_key_value_heads": 12,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"pretraining_tp": 1,
|
| 21 |
+
"rms_norm_eps": 1e-06,
|
| 22 |
+
"rope_scaling": null,
|
| 23 |
+
"rope_theta": 10000.0,
|
| 24 |
+
"tie_word_embeddings": false,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.41.0.dev0",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 32000
|
| 29 |
+
}
|
checkpoint-20000/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 1,
|
| 6 |
+
"transformers_version": "4.41.0.dev0"
|
| 7 |
+
}
|
checkpoint-20000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f548d3f50830b4785dda1f4538f7b5773c169769827be48f08631c46a71162f
|
| 3 |
+
size 272123144
|
checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ebd9aad19f6239a868ce7868c55a32316ce7f258db8495045466892dda3bb1a
|
| 3 |
+
size 544259743
|
checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d04e53e20ff2f98d1790a767178a0a2c6628c3c02f82f462c756ee89c4a2140
|
| 3 |
+
size 14244
|
checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d23c658a364caaa8e8f6fbe83d7a77bc1acc2b6786bd59a930d7c6edb5c62378
|
| 3 |
+
size 1000
|
checkpoint-20000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<unk>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-20000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-20000/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
checkpoint-20000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": true,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": true,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": true,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": true,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"bos_token": "<s>",
|
| 32 |
+
"clean_up_tokenization_spaces": false,
|
| 33 |
+
"eos_token": "</s>",
|
| 34 |
+
"legacy": true,
|
| 35 |
+
"model_max_length": 2048,
|
| 36 |
+
"pad_token": "<unk>",
|
| 37 |
+
"padding": "max_length",
|
| 38 |
+
"return_tensors": "pt",
|
| 39 |
+
"sp_model_kwargs": {},
|
| 40 |
+
"spaces_between_special_tokens": false,
|
| 41 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 42 |
+
"unk_token": "<unk>",
|
| 43 |
+
"use_default_system_prompt": false,
|
| 44 |
+
"use_fast": true
|
| 45 |
+
}
|
checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,493 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.644453259288642,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 20000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06611133148221605,
|
| 13 |
+
"grad_norm": 1.0232534408569336,
|
| 14 |
+
"learning_rate": 4.4080049369655294e-05,
|
| 15 |
+
"loss": 5.6838,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1322226629644321,
|
| 20 |
+
"grad_norm": 1.0166140794754028,
|
| 21 |
+
"learning_rate": 8.816009873931059e-05,
|
| 22 |
+
"loss": 3.8378,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1322226629644321,
|
| 27 |
+
"eval_accuracy": 0.4720710052887577,
|
| 28 |
+
"eval_loss": 3.6431853771209717,
|
| 29 |
+
"eval_runtime": 65.5904,
|
| 30 |
+
"eval_samples_per_second": 28.053,
|
| 31 |
+
"eval_steps_per_second": 1.174,
|
| 32 |
+
"step": 1000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.19833399444664815,
|
| 36 |
+
"grad_norm": 1.0467888116836548,
|
| 37 |
+
"learning_rate": 9.830315009952811e-05,
|
| 38 |
+
"loss": 3.3712,
|
| 39 |
+
"step": 1500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.2644453259288642,
|
| 43 |
+
"grad_norm": 1.1263777017593384,
|
| 44 |
+
"learning_rate": 9.59831475011252e-05,
|
| 45 |
+
"loss": 3.0922,
|
| 46 |
+
"step": 2000
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.2644453259288642,
|
| 50 |
+
"eval_accuracy": 0.5138609524011809,
|
| 51 |
+
"eval_loss": 3.076597213745117,
|
| 52 |
+
"eval_runtime": 64.6228,
|
| 53 |
+
"eval_samples_per_second": 28.473,
|
| 54 |
+
"eval_steps_per_second": 1.192,
|
| 55 |
+
"step": 2000
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.33055665741108026,
|
| 59 |
+
"grad_norm": 1.4438892602920532,
|
| 60 |
+
"learning_rate": 9.366314490272228e-05,
|
| 61 |
+
"loss": 2.9066,
|
| 62 |
+
"step": 2500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.3966679888932963,
|
| 66 |
+
"grad_norm": 1.3693314790725708,
|
| 67 |
+
"learning_rate": 9.134314230431938e-05,
|
| 68 |
+
"loss": 2.7993,
|
| 69 |
+
"step": 3000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.3966679888932963,
|
| 73 |
+
"eval_accuracy": 0.5319845054268176,
|
| 74 |
+
"eval_loss": 2.84745454788208,
|
| 75 |
+
"eval_runtime": 64.9029,
|
| 76 |
+
"eval_samples_per_second": 28.35,
|
| 77 |
+
"eval_steps_per_second": 1.186,
|
| 78 |
+
"step": 3000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.46277932037551234,
|
| 82 |
+
"grad_norm": 1.3279718160629272,
|
| 83 |
+
"learning_rate": 8.902313970591646e-05,
|
| 84 |
+
"loss": 2.7166,
|
| 85 |
+
"step": 3500
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.5288906518577284,
|
| 89 |
+
"grad_norm": 1.465155839920044,
|
| 90 |
+
"learning_rate": 8.670313710751356e-05,
|
| 91 |
+
"loss": 2.7115,
|
| 92 |
+
"step": 4000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.5288906518577284,
|
| 96 |
+
"eval_accuracy": 0.5392130052462777,
|
| 97 |
+
"eval_loss": 2.7528512477874756,
|
| 98 |
+
"eval_runtime": 65.039,
|
| 99 |
+
"eval_samples_per_second": 28.291,
|
| 100 |
+
"eval_steps_per_second": 1.184,
|
| 101 |
+
"step": 4000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.5950019833399445,
|
| 105 |
+
"grad_norm": 2.4618444442749023,
|
| 106 |
+
"learning_rate": 8.438313450911065e-05,
|
| 107 |
+
"loss": 2.644,
|
| 108 |
+
"step": 4500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.6611133148221605,
|
| 112 |
+
"grad_norm": 3.049086093902588,
|
| 113 |
+
"learning_rate": 8.206313191070773e-05,
|
| 114 |
+
"loss": 2.6702,
|
| 115 |
+
"step": 5000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.6611133148221605,
|
| 119 |
+
"eval_accuracy": 0.5420291625071685,
|
| 120 |
+
"eval_loss": 2.7150135040283203,
|
| 121 |
+
"eval_runtime": 64.9223,
|
| 122 |
+
"eval_samples_per_second": 28.342,
|
| 123 |
+
"eval_steps_per_second": 1.186,
|
| 124 |
+
"step": 5000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.7272246463043766,
|
| 128 |
+
"grad_norm": 3.927698850631714,
|
| 129 |
+
"learning_rate": 7.974312931230483e-05,
|
| 130 |
+
"loss": 2.6029,
|
| 131 |
+
"step": 5500
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.7933359777865926,
|
| 135 |
+
"grad_norm": 4.909026622772217,
|
| 136 |
+
"learning_rate": 7.742312671390191e-05,
|
| 137 |
+
"loss": 2.6484,
|
| 138 |
+
"step": 6000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.7933359777865926,
|
| 142 |
+
"eval_accuracy": 0.543187538497483,
|
| 143 |
+
"eval_loss": 2.696218729019165,
|
| 144 |
+
"eval_runtime": 64.8633,
|
| 145 |
+
"eval_samples_per_second": 28.367,
|
| 146 |
+
"eval_steps_per_second": 1.187,
|
| 147 |
+
"step": 6000
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.8594473092688086,
|
| 151 |
+
"grad_norm": 10.72818660736084,
|
| 152 |
+
"learning_rate": 7.510312411549901e-05,
|
| 153 |
+
"loss": 2.6474,
|
| 154 |
+
"step": 6500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.9255586407510247,
|
| 158 |
+
"grad_norm": 12.435935020446777,
|
| 159 |
+
"learning_rate": 7.278312151709609e-05,
|
| 160 |
+
"loss": 2.6419,
|
| 161 |
+
"step": 7000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.9255586407510247,
|
| 165 |
+
"eval_accuracy": 0.5387701514411334,
|
| 166 |
+
"eval_loss": 2.7223353385925293,
|
| 167 |
+
"eval_runtime": 68.3123,
|
| 168 |
+
"eval_samples_per_second": 26.935,
|
| 169 |
+
"eval_steps_per_second": 1.127,
|
| 170 |
+
"step": 7000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.9916699722332408,
|
| 174 |
+
"grad_norm": 15.605013847351074,
|
| 175 |
+
"learning_rate": 7.046311891869319e-05,
|
| 176 |
+
"loss": 2.6239,
|
| 177 |
+
"step": 7500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.0577813037154569,
|
| 181 |
+
"grad_norm": 55.199256896972656,
|
| 182 |
+
"learning_rate": 6.814311632029027e-05,
|
| 183 |
+
"loss": 2.5853,
|
| 184 |
+
"step": 8000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.0577813037154569,
|
| 188 |
+
"eval_accuracy": 0.5401743803232727,
|
| 189 |
+
"eval_loss": 2.7088677883148193,
|
| 190 |
+
"eval_runtime": 66.8855,
|
| 191 |
+
"eval_samples_per_second": 27.51,
|
| 192 |
+
"eval_steps_per_second": 1.151,
|
| 193 |
+
"step": 8000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.1238926351976728,
|
| 197 |
+
"grad_norm": 19.066770553588867,
|
| 198 |
+
"learning_rate": 6.582311372188736e-05,
|
| 199 |
+
"loss": 2.616,
|
| 200 |
+
"step": 8500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.190003966679889,
|
| 204 |
+
"grad_norm": 25.54907989501953,
|
| 205 |
+
"learning_rate": 6.350311112348446e-05,
|
| 206 |
+
"loss": 2.6009,
|
| 207 |
+
"step": 9000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.190003966679889,
|
| 211 |
+
"eval_accuracy": 0.5401903103162634,
|
| 212 |
+
"eval_loss": 2.703549861907959,
|
| 213 |
+
"eval_runtime": 65.0156,
|
| 214 |
+
"eval_samples_per_second": 28.301,
|
| 215 |
+
"eval_steps_per_second": 1.184,
|
| 216 |
+
"step": 9000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.256115298162105,
|
| 220 |
+
"grad_norm": 24.64689826965332,
|
| 221 |
+
"learning_rate": 6.118310852508154e-05,
|
| 222 |
+
"loss": 2.622,
|
| 223 |
+
"step": 9500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.322226629644321,
|
| 227 |
+
"grad_norm": 39.75895309448242,
|
| 228 |
+
"learning_rate": 5.886310592667864e-05,
|
| 229 |
+
"loss": 2.6347,
|
| 230 |
+
"step": 10000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.322226629644321,
|
| 234 |
+
"eval_accuracy": 0.5368293472950872,
|
| 235 |
+
"eval_loss": 2.7321841716766357,
|
| 236 |
+
"eval_runtime": 70.4109,
|
| 237 |
+
"eval_samples_per_second": 26.132,
|
| 238 |
+
"eval_steps_per_second": 1.094,
|
| 239 |
+
"step": 10000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.388337961126537,
|
| 243 |
+
"grad_norm": 63.85321044921875,
|
| 244 |
+
"learning_rate": 5.654310332827573e-05,
|
| 245 |
+
"loss": 2.634,
|
| 246 |
+
"step": 10500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.454449292608753,
|
| 250 |
+
"grad_norm": 38.93082046508789,
|
| 251 |
+
"learning_rate": 5.422310072987282e-05,
|
| 252 |
+
"loss": 2.7407,
|
| 253 |
+
"step": 11000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.454449292608753,
|
| 257 |
+
"eval_accuracy": 0.5244159002570039,
|
| 258 |
+
"eval_loss": 2.8357815742492676,
|
| 259 |
+
"eval_runtime": 64.8895,
|
| 260 |
+
"eval_samples_per_second": 28.356,
|
| 261 |
+
"eval_steps_per_second": 1.187,
|
| 262 |
+
"step": 11000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.5205606240909693,
|
| 266 |
+
"grad_norm": 95.31773376464844,
|
| 267 |
+
"learning_rate": 5.1903098131469904e-05,
|
| 268 |
+
"loss": 2.787,
|
| 269 |
+
"step": 11500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.5866719555731852,
|
| 273 |
+
"grad_norm": 388.4613952636719,
|
| 274 |
+
"learning_rate": 4.9583095533066995e-05,
|
| 275 |
+
"loss": 2.8981,
|
| 276 |
+
"step": 12000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.5866719555731852,
|
| 280 |
+
"eval_accuracy": 0.5073068222849982,
|
| 281 |
+
"eval_loss": 2.9791054725646973,
|
| 282 |
+
"eval_runtime": 65.179,
|
| 283 |
+
"eval_samples_per_second": 28.23,
|
| 284 |
+
"eval_steps_per_second": 1.181,
|
| 285 |
+
"step": 12000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.6527832870554013,
|
| 289 |
+
"grad_norm": 612.2838745117188,
|
| 290 |
+
"learning_rate": 4.7263092934664086e-05,
|
| 291 |
+
"loss": 2.9835,
|
| 292 |
+
"step": 12500
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.7188946185376173,
|
| 296 |
+
"grad_norm": 115.95877838134766,
|
| 297 |
+
"learning_rate": 4.4943090336261176e-05,
|
| 298 |
+
"loss": 3.1243,
|
| 299 |
+
"step": 13000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.7188946185376173,
|
| 303 |
+
"eval_accuracy": 0.4558367494318302,
|
| 304 |
+
"eval_loss": 3.454159736633301,
|
| 305 |
+
"eval_runtime": 64.9781,
|
| 306 |
+
"eval_samples_per_second": 28.317,
|
| 307 |
+
"eval_steps_per_second": 1.185,
|
| 308 |
+
"step": 13000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.7850059500198334,
|
| 312 |
+
"grad_norm": 152.87709045410156,
|
| 313 |
+
"learning_rate": 4.262308773785827e-05,
|
| 314 |
+
"loss": 3.2902,
|
| 315 |
+
"step": 13500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.8511172815020496,
|
| 319 |
+
"grad_norm": 235.900390625,
|
| 320 |
+
"learning_rate": 4.030308513945535e-05,
|
| 321 |
+
"loss": 3.2186,
|
| 322 |
+
"step": 14000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.8511172815020496,
|
| 326 |
+
"eval_accuracy": 0.4863910069879569,
|
| 327 |
+
"eval_loss": 3.138493061065674,
|
| 328 |
+
"eval_runtime": 65.7386,
|
| 329 |
+
"eval_samples_per_second": 27.99,
|
| 330 |
+
"eval_steps_per_second": 1.171,
|
| 331 |
+
"step": 14000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.9172286129842655,
|
| 335 |
+
"grad_norm": 22.522586822509766,
|
| 336 |
+
"learning_rate": 3.798308254105244e-05,
|
| 337 |
+
"loss": 3.0901,
|
| 338 |
+
"step": 14500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.9833399444664814,
|
| 342 |
+
"grad_norm": 186.2947235107422,
|
| 343 |
+
"learning_rate": 3.566307994264953e-05,
|
| 344 |
+
"loss": 2.9741,
|
| 345 |
+
"step": 15000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.9833399444664814,
|
| 349 |
+
"eval_accuracy": 0.4909350374885835,
|
| 350 |
+
"eval_loss": 3.0912961959838867,
|
| 351 |
+
"eval_runtime": 64.5786,
|
| 352 |
+
"eval_samples_per_second": 28.492,
|
| 353 |
+
"eval_steps_per_second": 1.192,
|
| 354 |
+
"step": 15000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.0494512759486976,
|
| 358 |
+
"grad_norm": 68.64359283447266,
|
| 359 |
+
"learning_rate": 3.3343077344246624e-05,
|
| 360 |
+
"loss": 2.8854,
|
| 361 |
+
"step": 15500
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.1155626074309137,
|
| 365 |
+
"grad_norm": 95.0623550415039,
|
| 366 |
+
"learning_rate": 3.1023074745843715e-05,
|
| 367 |
+
"loss": 2.8322,
|
| 368 |
+
"step": 16000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.1155626074309137,
|
| 372 |
+
"eval_accuracy": 0.5131106497313142,
|
| 373 |
+
"eval_loss": 2.899470329284668,
|
| 374 |
+
"eval_runtime": 64.8181,
|
| 375 |
+
"eval_samples_per_second": 28.387,
|
| 376 |
+
"eval_steps_per_second": 1.188,
|
| 377 |
+
"step": 16000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.18167393891313,
|
| 381 |
+
"grad_norm": 34.978458404541016,
|
| 382 |
+
"learning_rate": 2.8703072147440806e-05,
|
| 383 |
+
"loss": 2.8632,
|
| 384 |
+
"step": 16500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.2477852703953456,
|
| 388 |
+
"grad_norm": 94.10592651367188,
|
| 389 |
+
"learning_rate": 2.6383069549037897e-05,
|
| 390 |
+
"loss": 2.8482,
|
| 391 |
+
"step": 17000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.2477852703953456,
|
| 395 |
+
"eval_accuracy": 0.505882149911854,
|
| 396 |
+
"eval_loss": 2.944797992706299,
|
| 397 |
+
"eval_runtime": 64.8393,
|
| 398 |
+
"eval_samples_per_second": 28.378,
|
| 399 |
+
"eval_steps_per_second": 1.188,
|
| 400 |
+
"step": 17000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 2.3138966018775617,
|
| 404 |
+
"grad_norm": 68.33843994140625,
|
| 405 |
+
"learning_rate": 2.4063066950634984e-05,
|
| 406 |
+
"loss": 2.8747,
|
| 407 |
+
"step": 17500
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 2.380007933359778,
|
| 411 |
+
"grad_norm": 83.47583770751953,
|
| 412 |
+
"learning_rate": 2.1743064352232075e-05,
|
| 413 |
+
"loss": 2.8697,
|
| 414 |
+
"step": 18000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.380007933359778,
|
| 418 |
+
"eval_accuracy": 0.503569911429239,
|
| 419 |
+
"eval_loss": 2.9733448028564453,
|
| 420 |
+
"eval_runtime": 64.8844,
|
| 421 |
+
"eval_samples_per_second": 28.358,
|
| 422 |
+
"eval_steps_per_second": 1.187,
|
| 423 |
+
"step": 18000
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.446119264841994,
|
| 427 |
+
"grad_norm": 132.79673767089844,
|
| 428 |
+
"learning_rate": 1.9423061753829162e-05,
|
| 429 |
+
"loss": 2.8369,
|
| 430 |
+
"step": 18500
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.51223059632421,
|
| 434 |
+
"grad_norm": 189.53028869628906,
|
| 435 |
+
"learning_rate": 1.7103059155426253e-05,
|
| 436 |
+
"loss": 2.8289,
|
| 437 |
+
"step": 19000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.51223059632421,
|
| 441 |
+
"eval_accuracy": 0.5066844905588241,
|
| 442 |
+
"eval_loss": 2.94022536277771,
|
| 443 |
+
"eval_runtime": 65.2056,
|
| 444 |
+
"eval_samples_per_second": 28.218,
|
| 445 |
+
"eval_steps_per_second": 1.181,
|
| 446 |
+
"step": 19000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.578341927806426,
|
| 450 |
+
"grad_norm": 130.18309020996094,
|
| 451 |
+
"learning_rate": 1.4783056557023344e-05,
|
| 452 |
+
"loss": 2.8655,
|
| 453 |
+
"step": 19500
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 2.644453259288642,
|
| 457 |
+
"grad_norm": 136.07675170898438,
|
| 458 |
+
"learning_rate": 1.2463053958620433e-05,
|
| 459 |
+
"loss": 2.8319,
|
| 460 |
+
"step": 20000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 2.644453259288642,
|
| 464 |
+
"eval_accuracy": 0.506187209277628,
|
| 465 |
+
"eval_loss": 2.9402058124542236,
|
| 466 |
+
"eval_runtime": 65.0148,
|
| 467 |
+
"eval_samples_per_second": 28.301,
|
| 468 |
+
"eval_steps_per_second": 1.184,
|
| 469 |
+
"step": 20000
|
| 470 |
+
}
|
| 471 |
+
],
|
| 472 |
+
"logging_steps": 500,
|
| 473 |
+
"max_steps": 22689,
|
| 474 |
+
"num_input_tokens_seen": 0,
|
| 475 |
+
"num_train_epochs": 3,
|
| 476 |
+
"save_steps": 10000,
|
| 477 |
+
"stateful_callbacks": {
|
| 478 |
+
"TrainerControl": {
|
| 479 |
+
"args": {
|
| 480 |
+
"should_epoch_stop": false,
|
| 481 |
+
"should_evaluate": false,
|
| 482 |
+
"should_log": false,
|
| 483 |
+
"should_save": true,
|
| 484 |
+
"should_training_stop": false
|
| 485 |
+
},
|
| 486 |
+
"attributes": {}
|
| 487 |
+
}
|
| 488 |
+
},
|
| 489 |
+
"total_flos": 1.2813992666018611e+17,
|
| 490 |
+
"train_batch_size": 12,
|
| 491 |
+
"trial_name": null,
|
| 492 |
+
"trial_params": null
|
| 493 |
+
}
|
checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b1788df2e878fccca08354d00fe3ff798a83f0940758c72244a04260d9cb1f0
|
| 3 |
+
size 5176
|