teacher fwd distill
Browse files- config.json +26 -0
- generation_config.json +7 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +24 -0
- tokenizer.model +3 -0
- tokenizer_config.json +35 -0
- trainer_state.json +673 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "JackFram/llama-160m",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"hidden_act": "silu",
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"max_position_embeddings": 2048,
|
| 13 |
+
"model_type": "llama",
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 12,
|
| 16 |
+
"num_key_value_heads": 12,
|
| 17 |
+
"pad_token_id": 1,
|
| 18 |
+
"pretraining_tp": 1,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"tie_word_embeddings": false,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.31.0",
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 32000
|
| 26 |
+
}
|
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": "4.31.0"
|
| 7 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d8dcbb8e5a58065555c19b7113c828aa705d33aeccfa5b23e8a6d2ee7c10497
|
| 3 |
+
size 649710457
|
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 |
+
}
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"bos_token": {
|
| 5 |
+
"__type": "AddedToken",
|
| 6 |
+
"content": "<s>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"clean_up_tokenization_spaces": false,
|
| 13 |
+
"eos_token": {
|
| 14 |
+
"__type": "AddedToken",
|
| 15 |
+
"content": "</s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": true,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false
|
| 20 |
+
},
|
| 21 |
+
"legacy": true,
|
| 22 |
+
"model_max_length": 2048,
|
| 23 |
+
"pad_token": null,
|
| 24 |
+
"padding_side": "right",
|
| 25 |
+
"sp_model_kwargs": {},
|
| 26 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 27 |
+
"unk_token": {
|
| 28 |
+
"__type": "AddedToken",
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false
|
| 34 |
+
}
|
| 35 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,673 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9748571428571429,
|
| 5 |
+
"global_step": 108,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 5e-06,
|
| 13 |
+
"loss": 5.2929,
|
| 14 |
+
"step": 1
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 1e-05,
|
| 19 |
+
"loss": 5.1707,
|
| 20 |
+
"step": 2
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.05,
|
| 24 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 25 |
+
"loss": 4.86,
|
| 26 |
+
"step": 3
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.07,
|
| 30 |
+
"learning_rate": 2e-05,
|
| 31 |
+
"loss": 4.909,
|
| 32 |
+
"step": 4
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.09,
|
| 36 |
+
"learning_rate": 1.9995437844895337e-05,
|
| 37 |
+
"loss": 4.4727,
|
| 38 |
+
"step": 5
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.11,
|
| 42 |
+
"learning_rate": 1.9981755542233175e-05,
|
| 43 |
+
"loss": 4.1336,
|
| 44 |
+
"step": 6
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.13,
|
| 48 |
+
"learning_rate": 1.995896557617091e-05,
|
| 49 |
+
"loss": 3.5284,
|
| 50 |
+
"step": 7
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.15,
|
| 54 |
+
"learning_rate": 1.992708874098054e-05,
|
| 55 |
+
"loss": 3.1435,
|
| 56 |
+
"step": 8
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.16,
|
| 60 |
+
"learning_rate": 1.9886154122075344e-05,
|
| 61 |
+
"loss": 2.9837,
|
| 62 |
+
"step": 9
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.18,
|
| 66 |
+
"learning_rate": 1.983619906947144e-05,
|
| 67 |
+
"loss": 2.6885,
|
| 68 |
+
"step": 10
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.2,
|
| 72 |
+
"learning_rate": 1.977726916370847e-05,
|
| 73 |
+
"loss": 2.6187,
|
| 74 |
+
"step": 11
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.22,
|
| 78 |
+
"learning_rate": 1.9709418174260523e-05,
|
| 79 |
+
"loss": 2.5325,
|
| 80 |
+
"step": 12
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24,
|
| 84 |
+
"learning_rate": 1.9632708010475166e-05,
|
| 85 |
+
"loss": 2.2144,
|
| 86 |
+
"step": 13
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26,
|
| 90 |
+
"learning_rate": 1.954720866508546e-05,
|
| 91 |
+
"loss": 2.1453,
|
| 92 |
+
"step": 14
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.27,
|
| 96 |
+
"learning_rate": 1.9452998150346403e-05,
|
| 97 |
+
"loss": 2.0979,
|
| 98 |
+
"step": 15
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.29,
|
| 102 |
+
"learning_rate": 1.9350162426854152e-05,
|
| 103 |
+
"loss": 2.1127,
|
| 104 |
+
"step": 16
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.31,
|
| 108 |
+
"learning_rate": 1.9238795325112867e-05,
|
| 109 |
+
"loss": 1.9363,
|
| 110 |
+
"step": 17
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.33,
|
| 114 |
+
"learning_rate": 1.91189984599209e-05,
|
| 115 |
+
"loss": 1.9072,
|
| 116 |
+
"step": 18
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.35,
|
| 120 |
+
"learning_rate": 1.899088113765426e-05,
|
| 121 |
+
"loss": 1.8657,
|
| 122 |
+
"step": 19
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.37,
|
| 126 |
+
"learning_rate": 1.8854560256532098e-05,
|
| 127 |
+
"loss": 1.9458,
|
| 128 |
+
"step": 20
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.38,
|
| 132 |
+
"learning_rate": 1.8710160199955158e-05,
|
| 133 |
+
"loss": 1.8255,
|
| 134 |
+
"step": 21
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.4,
|
| 138 |
+
"learning_rate": 1.8557812723014476e-05,
|
| 139 |
+
"loss": 1.9809,
|
| 140 |
+
"step": 22
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.42,
|
| 144 |
+
"learning_rate": 1.8397656832273982e-05,
|
| 145 |
+
"loss": 1.4541,
|
| 146 |
+
"step": 23
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.44,
|
| 150 |
+
"learning_rate": 1.8229838658936566e-05,
|
| 151 |
+
"loss": 1.496,
|
| 152 |
+
"step": 24
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.46,
|
| 156 |
+
"learning_rate": 1.805451132550946e-05,
|
| 157 |
+
"loss": 1.6486,
|
| 158 |
+
"step": 25
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.48,
|
| 162 |
+
"learning_rate": 1.7871834806090502e-05,
|
| 163 |
+
"loss": 1.4413,
|
| 164 |
+
"step": 26
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.49,
|
| 168 |
+
"learning_rate": 1.7681975780402807e-05,
|
| 169 |
+
"loss": 1.4627,
|
| 170 |
+
"step": 27
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.51,
|
| 174 |
+
"learning_rate": 1.7485107481711014e-05,
|
| 175 |
+
"loss": 1.548,
|
| 176 |
+
"step": 28
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.53,
|
| 180 |
+
"learning_rate": 1.7281409538757886e-05,
|
| 181 |
+
"loss": 1.3278,
|
| 182 |
+
"step": 29
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.55,
|
| 186 |
+
"learning_rate": 1.7071067811865477e-05,
|
| 187 |
+
"loss": 1.4126,
|
| 188 |
+
"step": 30
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.57,
|
| 192 |
+
"learning_rate": 1.68542742233504e-05,
|
| 193 |
+
"loss": 1.6469,
|
| 194 |
+
"step": 31
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.59,
|
| 198 |
+
"learning_rate": 1.6631226582407954e-05,
|
| 199 |
+
"loss": 1.2131,
|
| 200 |
+
"step": 32
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.6,
|
| 204 |
+
"learning_rate": 1.640212840462488e-05,
|
| 205 |
+
"loss": 1.4731,
|
| 206 |
+
"step": 33
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.62,
|
| 210 |
+
"learning_rate": 1.6167188726285433e-05,
|
| 211 |
+
"loss": 1.2919,
|
| 212 |
+
"step": 34
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.64,
|
| 216 |
+
"learning_rate": 1.592662191364017e-05,
|
| 217 |
+
"loss": 1.3771,
|
| 218 |
+
"step": 35
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.66,
|
| 222 |
+
"learning_rate": 1.568064746731156e-05,
|
| 223 |
+
"loss": 1.3238,
|
| 224 |
+
"step": 36
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.68,
|
| 228 |
+
"learning_rate": 1.542948982201479e-05,
|
| 229 |
+
"loss": 1.2339,
|
| 230 |
+
"step": 37
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.69,
|
| 234 |
+
"learning_rate": 1.5173378141776569e-05,
|
| 235 |
+
"loss": 1.2227,
|
| 236 |
+
"step": 38
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.71,
|
| 240 |
+
"learning_rate": 1.4912546110838775e-05,
|
| 241 |
+
"loss": 1.1856,
|
| 242 |
+
"step": 39
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.73,
|
| 246 |
+
"learning_rate": 1.4647231720437687e-05,
|
| 247 |
+
"loss": 1.2679,
|
| 248 |
+
"step": 40
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.75,
|
| 252 |
+
"learning_rate": 1.4377677051653404e-05,
|
| 253 |
+
"loss": 1.1049,
|
| 254 |
+
"step": 41
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.77,
|
| 258 |
+
"learning_rate": 1.410412805452757e-05,
|
| 259 |
+
"loss": 1.0227,
|
| 260 |
+
"step": 42
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.79,
|
| 264 |
+
"learning_rate": 1.3826834323650899e-05,
|
| 265 |
+
"loss": 1.2718,
|
| 266 |
+
"step": 43
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.8,
|
| 270 |
+
"learning_rate": 1.3546048870425356e-05,
|
| 271 |
+
"loss": 0.9942,
|
| 272 |
+
"step": 44
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.82,
|
| 276 |
+
"learning_rate": 1.3262027892208696e-05,
|
| 277 |
+
"loss": 1.2092,
|
| 278 |
+
"step": 45
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.84,
|
| 282 |
+
"learning_rate": 1.297503053855203e-05,
|
| 283 |
+
"loss": 1.0914,
|
| 284 |
+
"step": 46
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.86,
|
| 288 |
+
"learning_rate": 1.2685318674743769e-05,
|
| 289 |
+
"loss": 1.104,
|
| 290 |
+
"step": 47
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.88,
|
| 294 |
+
"learning_rate": 1.2393156642875579e-05,
|
| 295 |
+
"loss": 1.011,
|
| 296 |
+
"step": 48
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.9,
|
| 300 |
+
"learning_rate": 1.2098811020648475e-05,
|
| 301 |
+
"loss": 1.4294,
|
| 302 |
+
"step": 49
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.91,
|
| 306 |
+
"learning_rate": 1.180255037813906e-05,
|
| 307 |
+
"loss": 1.1338,
|
| 308 |
+
"step": 50
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.93,
|
| 312 |
+
"learning_rate": 1.1504645032747832e-05,
|
| 313 |
+
"loss": 1.154,
|
| 314 |
+
"step": 51
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.95,
|
| 318 |
+
"learning_rate": 1.1205366802553231e-05,
|
| 319 |
+
"loss": 1.0417,
|
| 320 |
+
"step": 52
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.97,
|
| 324 |
+
"learning_rate": 1.090498875829638e-05,
|
| 325 |
+
"loss": 0.89,
|
| 326 |
+
"step": 53
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.99,
|
| 330 |
+
"learning_rate": 1.0603784974222862e-05,
|
| 331 |
+
"loss": 0.8541,
|
| 332 |
+
"step": 54
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.01,
|
| 336 |
+
"learning_rate": 1.030203027800889e-05,
|
| 337 |
+
"loss": 1.0875,
|
| 338 |
+
"step": 55
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.02,
|
| 342 |
+
"learning_rate": 1e-05,
|
| 343 |
+
"loss": 1.1117,
|
| 344 |
+
"step": 56
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.04,
|
| 348 |
+
"learning_rate": 9.697969721991114e-06,
|
| 349 |
+
"loss": 1.0332,
|
| 350 |
+
"step": 57
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.06,
|
| 354 |
+
"learning_rate": 9.39621502577714e-06,
|
| 355 |
+
"loss": 0.9024,
|
| 356 |
+
"step": 58
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.08,
|
| 360 |
+
"learning_rate": 9.095011241703623e-06,
|
| 361 |
+
"loss": 0.8321,
|
| 362 |
+
"step": 59
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.1,
|
| 366 |
+
"learning_rate": 8.79463319744677e-06,
|
| 367 |
+
"loss": 0.817,
|
| 368 |
+
"step": 60
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.12,
|
| 372 |
+
"learning_rate": 8.49535496725217e-06,
|
| 373 |
+
"loss": 0.8457,
|
| 374 |
+
"step": 61
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.13,
|
| 378 |
+
"learning_rate": 8.197449621860944e-06,
|
| 379 |
+
"loss": 0.9017,
|
| 380 |
+
"step": 62
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.15,
|
| 384 |
+
"learning_rate": 7.901188979351527e-06,
|
| 385 |
+
"loss": 1.0759,
|
| 386 |
+
"step": 63
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.17,
|
| 390 |
+
"learning_rate": 7.606843357124426e-06,
|
| 391 |
+
"loss": 1.011,
|
| 392 |
+
"step": 64
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.19,
|
| 396 |
+
"learning_rate": 7.314681325256232e-06,
|
| 397 |
+
"loss": 0.8737,
|
| 398 |
+
"step": 65
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.21,
|
| 402 |
+
"learning_rate": 7.024969461447973e-06,
|
| 403 |
+
"loss": 0.9498,
|
| 404 |
+
"step": 66
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.23,
|
| 408 |
+
"learning_rate": 6.7379721077913095e-06,
|
| 409 |
+
"loss": 1.0125,
|
| 410 |
+
"step": 67
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.24,
|
| 414 |
+
"learning_rate": 6.453951129574644e-06,
|
| 415 |
+
"loss": 0.7371,
|
| 416 |
+
"step": 68
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.26,
|
| 420 |
+
"learning_rate": 6.173165676349103e-06,
|
| 421 |
+
"loss": 0.8445,
|
| 422 |
+
"step": 69
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.28,
|
| 426 |
+
"learning_rate": 5.895871945472434e-06,
|
| 427 |
+
"loss": 1.0347,
|
| 428 |
+
"step": 70
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.3,
|
| 432 |
+
"learning_rate": 5.622322948346595e-06,
|
| 433 |
+
"loss": 1.0074,
|
| 434 |
+
"step": 71
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.32,
|
| 438 |
+
"learning_rate": 5.352768279562315e-06,
|
| 439 |
+
"loss": 0.887,
|
| 440 |
+
"step": 72
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.33,
|
| 444 |
+
"learning_rate": 5.087453889161229e-06,
|
| 445 |
+
"loss": 0.9886,
|
| 446 |
+
"step": 73
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.35,
|
| 450 |
+
"learning_rate": 4.826621858223431e-06,
|
| 451 |
+
"loss": 0.9414,
|
| 452 |
+
"step": 74
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.37,
|
| 456 |
+
"learning_rate": 4.570510177985213e-06,
|
| 457 |
+
"loss": 1.098,
|
| 458 |
+
"step": 75
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.39,
|
| 462 |
+
"learning_rate": 4.319352532688444e-06,
|
| 463 |
+
"loss": 1.0022,
|
| 464 |
+
"step": 76
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.41,
|
| 468 |
+
"learning_rate": 4.073378086359834e-06,
|
| 469 |
+
"loss": 0.9925,
|
| 470 |
+
"step": 77
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.43,
|
| 474 |
+
"learning_rate": 3.832811273714569e-06,
|
| 475 |
+
"loss": 0.7212,
|
| 476 |
+
"step": 78
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.44,
|
| 480 |
+
"learning_rate": 3.5978715953751207e-06,
|
| 481 |
+
"loss": 0.7285,
|
| 482 |
+
"step": 79
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.46,
|
| 486 |
+
"learning_rate": 3.3687734175920505e-06,
|
| 487 |
+
"loss": 0.8824,
|
| 488 |
+
"step": 80
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.48,
|
| 492 |
+
"learning_rate": 3.145725776649602e-06,
|
| 493 |
+
"loss": 0.8606,
|
| 494 |
+
"step": 81
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.5,
|
| 498 |
+
"learning_rate": 2.9289321881345257e-06,
|
| 499 |
+
"loss": 0.7898,
|
| 500 |
+
"step": 82
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.52,
|
| 504 |
+
"learning_rate": 2.7185904612421177e-06,
|
| 505 |
+
"loss": 0.913,
|
| 506 |
+
"step": 83
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.54,
|
| 510 |
+
"learning_rate": 2.514892518288988e-06,
|
| 511 |
+
"loss": 0.9278,
|
| 512 |
+
"step": 84
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.55,
|
| 516 |
+
"learning_rate": 2.318024219597196e-06,
|
| 517 |
+
"loss": 0.9006,
|
| 518 |
+
"step": 85
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.57,
|
| 522 |
+
"learning_rate": 2.1281651939094996e-06,
|
| 523 |
+
"loss": 0.9705,
|
| 524 |
+
"step": 86
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.59,
|
| 528 |
+
"learning_rate": 1.94548867449054e-06,
|
| 529 |
+
"loss": 0.848,
|
| 530 |
+
"step": 87
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.61,
|
| 534 |
+
"learning_rate": 1.7701613410634367e-06,
|
| 535 |
+
"loss": 0.947,
|
| 536 |
+
"step": 88
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.63,
|
| 540 |
+
"learning_rate": 1.6023431677260215e-06,
|
| 541 |
+
"loss": 0.8865,
|
| 542 |
+
"step": 89
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.65,
|
| 546 |
+
"learning_rate": 1.4421872769855262e-06,
|
| 547 |
+
"loss": 0.9392,
|
| 548 |
+
"step": 90
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.66,
|
| 552 |
+
"learning_rate": 1.2898398000448441e-06,
|
| 553 |
+
"loss": 0.85,
|
| 554 |
+
"step": 91
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.68,
|
| 558 |
+
"learning_rate": 1.1454397434679022e-06,
|
| 559 |
+
"loss": 0.7694,
|
| 560 |
+
"step": 92
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.7,
|
| 564 |
+
"learning_rate": 1.0091188623457415e-06,
|
| 565 |
+
"loss": 0.8638,
|
| 566 |
+
"step": 93
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.72,
|
| 570 |
+
"learning_rate": 8.810015400790994e-07,
|
| 571 |
+
"loss": 0.8655,
|
| 572 |
+
"step": 94
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.74,
|
| 576 |
+
"learning_rate": 7.612046748871327e-07,
|
| 577 |
+
"loss": 0.8398,
|
| 578 |
+
"step": 95
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.76,
|
| 582 |
+
"learning_rate": 6.498375731458529e-07,
|
| 583 |
+
"loss": 0.7379,
|
| 584 |
+
"step": 96
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.77,
|
| 588 |
+
"learning_rate": 5.470018496535967e-07,
|
| 589 |
+
"loss": 0.8092,
|
| 590 |
+
"step": 97
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.79,
|
| 594 |
+
"learning_rate": 4.5279133491454406e-07,
|
| 595 |
+
"loss": 0.8303,
|
| 596 |
+
"step": 98
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.81,
|
| 600 |
+
"learning_rate": 3.6729198952483725e-07,
|
| 601 |
+
"loss": 0.7979,
|
| 602 |
+
"step": 99
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.83,
|
| 606 |
+
"learning_rate": 2.905818257394799e-07,
|
| 607 |
+
"loss": 0.834,
|
| 608 |
+
"step": 100
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.85,
|
| 612 |
+
"learning_rate": 2.2273083629153148e-07,
|
| 613 |
+
"loss": 0.8447,
|
| 614 |
+
"step": 101
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.87,
|
| 618 |
+
"learning_rate": 1.6380093052856482e-07,
|
| 619 |
+
"loss": 0.8207,
|
| 620 |
+
"step": 102
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.88,
|
| 624 |
+
"learning_rate": 1.1384587792465873e-07,
|
| 625 |
+
"loss": 0.8542,
|
| 626 |
+
"step": 103
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.9,
|
| 630 |
+
"learning_rate": 7.291125901946027e-08,
|
| 631 |
+
"loss": 1.091,
|
| 632 |
+
"step": 104
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.92,
|
| 636 |
+
"learning_rate": 4.103442382909051e-08,
|
| 637 |
+
"loss": 0.9051,
|
| 638 |
+
"step": 105
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.94,
|
| 642 |
+
"learning_rate": 1.824445776682504e-08,
|
| 643 |
+
"loss": 0.9512,
|
| 644 |
+
"step": 106
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.96,
|
| 648 |
+
"learning_rate": 4.562155104665955e-09,
|
| 649 |
+
"loss": 0.6725,
|
| 650 |
+
"step": 107
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.97,
|
| 654 |
+
"learning_rate": 0.0,
|
| 655 |
+
"loss": 0.7599,
|
| 656 |
+
"step": 108
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.97,
|
| 660 |
+
"step": 108,
|
| 661 |
+
"total_flos": 2.3415025142071296e+16,
|
| 662 |
+
"train_loss": 1.4361559649308522,
|
| 663 |
+
"train_runtime": 6959.0247,
|
| 664 |
+
"train_samples_per_second": 2.012,
|
| 665 |
+
"train_steps_per_second": 0.016
|
| 666 |
+
}
|
| 667 |
+
],
|
| 668 |
+
"max_steps": 108,
|
| 669 |
+
"num_train_epochs": 2,
|
| 670 |
+
"total_flos": 2.3415025142071296e+16,
|
| 671 |
+
"trial_name": null,
|
| 672 |
+
"trial_params": null
|
| 673 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5ed9437503d74cbd04a3a06445c7492650fea881f0064978ff9e1ac66bfb633
|
| 3 |
+
size 4219
|