Upload folder using huggingface_hub
Browse files- added_tokens.json +3 -0
- config.json +31 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +51 -0
- tokenizer.json +0 -0
- tokenizer_config.json +57 -0
- trainer_state.json +548 -0
- training_args.bin +3 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<mask>": 50000
|
| 3 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/avijit/playground/sdoh/RoBERTa-base-PM-M3-Voc-distill-align",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForPromptBasedMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"label_no": 689,
|
| 17 |
+
"label_yes": 25597,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"mask_token_id": 50000,
|
| 20 |
+
"max_position_embeddings": 514,
|
| 21 |
+
"model_type": "roberta",
|
| 22 |
+
"num_attention_heads": 12,
|
| 23 |
+
"num_hidden_layers": 12,
|
| 24 |
+
"pad_token_id": 1,
|
| 25 |
+
"position_embedding_type": "absolute",
|
| 26 |
+
"torch_dtype": "float32",
|
| 27 |
+
"transformers_version": "4.39.1",
|
| 28 |
+
"type_vocab_size": 1,
|
| 29 |
+
"use_cache": true,
|
| 30 |
+
"vocab_size": 50008
|
| 31 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f95859c07074c7367e34fc23abe44741a3eb9498ff1c1bdfaa24b99b9d3a860
|
| 3 |
+
size 498023416
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1bc5a19cb6f7f978326969d77a39ae7737a848b3a8e06d80f66728490d8b952
|
| 3 |
+
size 996167290
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2415f42972bda2f4abbefa71ee8b14cb2093f8afa0f03ba03d0c03d490cbfff5
|
| 3 |
+
size 14180
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88e186000a6a6fb46572875c44f86929266570aa0e96bf9bf3263e05e25b5b0d
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": true,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": true,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"50000": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"bos_token": "<s>",
|
| 46 |
+
"clean_up_tokenization_spaces": true,
|
| 47 |
+
"cls_token": "<s>",
|
| 48 |
+
"eos_token": "</s>",
|
| 49 |
+
"errors": "replace",
|
| 50 |
+
"mask_token": "<mask>",
|
| 51 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 52 |
+
"pad_token": "<pad>",
|
| 53 |
+
"sep_token": "</s>",
|
| 54 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 55 |
+
"trim_offsets": true,
|
| 56 |
+
"unk_token": "<unk>"
|
| 57 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,548 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9235158618894997,
|
| 3 |
+
"best_model_checkpoint": "./saved_models/cliroberta_prompt_sbdh_gpt4_v2_0/checkpoint-408",
|
| 4 |
+
"epoch": 17.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 408,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 6.645392417907715,
|
| 14 |
+
"learning_rate": 2.5e-06,
|
| 15 |
+
"loss": 1.8965,
|
| 16 |
+
"step": 24
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_acc_macro": 0.07351644449088429,
|
| 21 |
+
"eval_acc_micro": 0.16844170403586972,
|
| 22 |
+
"eval_auc_macro": 0.5519856670284093,
|
| 23 |
+
"eval_auc_micro": 0.6260212922777759,
|
| 24 |
+
"eval_f1_at_5": 0.25157489791625554,
|
| 25 |
+
"eval_f1_at_8": 0.2244759753473909,
|
| 26 |
+
"eval_f1_macro": 0.12396975795306268,
|
| 27 |
+
"eval_f1_micro": 0.2883185416166808,
|
| 28 |
+
"eval_loss": 0.3488869071006775,
|
| 29 |
+
"eval_prec_at_5": 0.16232876712328767,
|
| 30 |
+
"eval_prec_at_8": 0.13227739726027396,
|
| 31 |
+
"eval_prec_macro": 0.1633962475788553,
|
| 32 |
+
"eval_prec_micro": 0.20283496456293612,
|
| 33 |
+
"eval_rec_at_5": 0.5587899543378996,
|
| 34 |
+
"eval_rec_at_8": 0.7408675799086758,
|
| 35 |
+
"eval_rec_macro": 0.3086744639374022,
|
| 36 |
+
"eval_rec_micro": 0.4983416252072555,
|
| 37 |
+
"eval_runtime": 4.7343,
|
| 38 |
+
"eval_samples_per_second": 185.032,
|
| 39 |
+
"eval_steps_per_second": 23.235,
|
| 40 |
+
"step": 24
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 2.0,
|
| 44 |
+
"grad_norm": 6.26666784286499,
|
| 45 |
+
"learning_rate": 5e-06,
|
| 46 |
+
"loss": 0.3531,
|
| 47 |
+
"step": 48
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.0,
|
| 51 |
+
"eval_acc_macro": 0.11095425738710896,
|
| 52 |
+
"eval_acc_micro": 0.23921319796952797,
|
| 53 |
+
"eval_auc_macro": 0.8260589461583061,
|
| 54 |
+
"eval_auc_micro": 0.7608045184921441,
|
| 55 |
+
"eval_f1_at_5": 0.3274833326039068,
|
| 56 |
+
"eval_f1_at_8": 0.26273329679353813,
|
| 57 |
+
"eval_f1_macro": 0.16550844622182223,
|
| 58 |
+
"eval_f1_micro": 0.3860727086533143,
|
| 59 |
+
"eval_loss": 0.3172319531440735,
|
| 60 |
+
"eval_prec_at_5": 0.2100456621004566,
|
| 61 |
+
"eval_prec_at_8": 0.1545376712328767,
|
| 62 |
+
"eval_prec_macro": 0.36732877721657925,
|
| 63 |
+
"eval_prec_micro": 0.5046854082997986,
|
| 64 |
+
"eval_rec_at_5": 0.7427701674277017,
|
| 65 |
+
"eval_rec_at_8": 0.8761415525114156,
|
| 66 |
+
"eval_rec_macro": 0.1808019430800684,
|
| 67 |
+
"eval_rec_micro": 0.312603648424518,
|
| 68 |
+
"eval_runtime": 6.9454,
|
| 69 |
+
"eval_samples_per_second": 126.127,
|
| 70 |
+
"eval_steps_per_second": 15.838,
|
| 71 |
+
"step": 48
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 3.0,
|
| 75 |
+
"grad_norm": 1.4328076839447021,
|
| 76 |
+
"learning_rate": 7.500000000000001e-06,
|
| 77 |
+
"loss": 0.2409,
|
| 78 |
+
"step": 72
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.0,
|
| 82 |
+
"eval_acc_macro": 0.6542871174273012,
|
| 83 |
+
"eval_acc_micro": 0.6627906976743733,
|
| 84 |
+
"eval_auc_macro": 0.9683337499122936,
|
| 85 |
+
"eval_auc_micro": 0.9677785939562165,
|
| 86 |
+
"eval_f1_at_5": 0.4172108144155159,
|
| 87 |
+
"eval_f1_at_8": 0.29078564566802617,
|
| 88 |
+
"eval_f1_macro": 0.7752458106791931,
|
| 89 |
+
"eval_f1_micro": 0.7972027972027316,
|
| 90 |
+
"eval_loss": 0.1295301616191864,
|
| 91 |
+
"eval_prec_at_5": 0.2671232876712329,
|
| 92 |
+
"eval_prec_at_8": 0.1709474885844749,
|
| 93 |
+
"eval_prec_macro": 0.7878146103091771,
|
| 94 |
+
"eval_prec_micro": 0.7910204081632007,
|
| 95 |
+
"eval_rec_at_5": 0.9522450532724505,
|
| 96 |
+
"eval_rec_at_8": 0.9726027397260274,
|
| 97 |
+
"eval_rec_macro": 0.7778465051663405,
|
| 98 |
+
"eval_rec_micro": 0.80348258706461,
|
| 99 |
+
"eval_runtime": 7.0065,
|
| 100 |
+
"eval_samples_per_second": 125.028,
|
| 101 |
+
"eval_steps_per_second": 15.7,
|
| 102 |
+
"step": 72
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 4.0,
|
| 106 |
+
"grad_norm": 2.6871750354766846,
|
| 107 |
+
"learning_rate": 1e-05,
|
| 108 |
+
"loss": 0.1262,
|
| 109 |
+
"step": 96
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 4.0,
|
| 113 |
+
"eval_acc_macro": 0.7556596335774451,
|
| 114 |
+
"eval_acc_micro": 0.7631578947367863,
|
| 115 |
+
"eval_auc_macro": 0.9859339445130657,
|
| 116 |
+
"eval_auc_micro": 0.9878248630762655,
|
| 117 |
+
"eval_f1_at_5": 0.4281991387591176,
|
| 118 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 119 |
+
"eval_f1_macro": 0.8563120619008121,
|
| 120 |
+
"eval_f1_micro": 0.865671641790973,
|
| 121 |
+
"eval_loss": 0.086879663169384,
|
| 122 |
+
"eval_prec_at_5": 0.2742009132420092,
|
| 123 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 124 |
+
"eval_prec_macro": 0.8563984905449645,
|
| 125 |
+
"eval_prec_micro": 0.865671641790973,
|
| 126 |
+
"eval_rec_at_5": 0.9767884322678843,
|
| 127 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 128 |
+
"eval_rec_macro": 0.8591516601823509,
|
| 129 |
+
"eval_rec_micro": 0.865671641790973,
|
| 130 |
+
"eval_runtime": 7.0436,
|
| 131 |
+
"eval_samples_per_second": 124.369,
|
| 132 |
+
"eval_steps_per_second": 15.617,
|
| 133 |
+
"step": 96
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 5.0,
|
| 137 |
+
"grad_norm": 1.409122347831726,
|
| 138 |
+
"learning_rate": 9.722222222222223e-06,
|
| 139 |
+
"loss": 0.0916,
|
| 140 |
+
"step": 120
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 5.0,
|
| 144 |
+
"eval_acc_macro": 0.8023972093976516,
|
| 145 |
+
"eval_acc_micro": 0.8076079005120111,
|
| 146 |
+
"eval_auc_macro": 0.9910191597807554,
|
| 147 |
+
"eval_auc_micro": 0.99171908563779,
|
| 148 |
+
"eval_f1_at_5": 0.4281442700601774,
|
| 149 |
+
"eval_f1_at_8": 0.29274329072668936,
|
| 150 |
+
"eval_f1_macro": 0.8879651934832161,
|
| 151 |
+
"eval_f1_micro": 0.893565358154521,
|
| 152 |
+
"eval_loss": 0.06939995288848877,
|
| 153 |
+
"eval_prec_at_5": 0.2742009132420092,
|
| 154 |
+
"eval_prec_at_8": 0.1720890410958904,
|
| 155 |
+
"eval_prec_macro": 0.8664154958281926,
|
| 156 |
+
"eval_prec_micro": 0.8727272727272037,
|
| 157 |
+
"eval_rec_at_5": 0.9762176560121765,
|
| 158 |
+
"eval_rec_at_8": 0.9794520547945206,
|
| 159 |
+
"eval_rec_macro": 0.9116566511084603,
|
| 160 |
+
"eval_rec_micro": 0.9154228855720634,
|
| 161 |
+
"eval_runtime": 7.006,
|
| 162 |
+
"eval_samples_per_second": 125.035,
|
| 163 |
+
"eval_steps_per_second": 15.701,
|
| 164 |
+
"step": 120
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 6.0,
|
| 168 |
+
"grad_norm": 1.1020153760910034,
|
| 169 |
+
"learning_rate": 9.444444444444445e-06,
|
| 170 |
+
"loss": 0.0785,
|
| 171 |
+
"step": 144
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 6.0,
|
| 175 |
+
"eval_acc_macro": 0.8155311593684006,
|
| 176 |
+
"eval_acc_micro": 0.8230421686746368,
|
| 177 |
+
"eval_auc_macro": 0.9922239037332781,
|
| 178 |
+
"eval_auc_micro": 0.9932765964575004,
|
| 179 |
+
"eval_f1_at_5": 0.4288655829720509,
|
| 180 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 181 |
+
"eval_f1_macro": 0.8957459298097094,
|
| 182 |
+
"eval_f1_micro": 0.9029326724493264,
|
| 183 |
+
"eval_loss": 0.06519585102796555,
|
| 184 |
+
"eval_prec_at_5": 0.27465753424657535,
|
| 185 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 186 |
+
"eval_prec_macro": 0.8920600176459569,
|
| 187 |
+
"eval_prec_micro": 0.8995884773661811,
|
| 188 |
+
"eval_rec_at_5": 0.9779299847792998,
|
| 189 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 190 |
+
"eval_rec_macro": 0.900216189836664,
|
| 191 |
+
"eval_rec_micro": 0.9063018242121967,
|
| 192 |
+
"eval_runtime": 7.0038,
|
| 193 |
+
"eval_samples_per_second": 125.075,
|
| 194 |
+
"eval_steps_per_second": 15.706,
|
| 195 |
+
"step": 144
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 7.0,
|
| 199 |
+
"grad_norm": 1.087604284286499,
|
| 200 |
+
"learning_rate": 9.166666666666666e-06,
|
| 201 |
+
"loss": 0.0692,
|
| 202 |
+
"step": 168
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 7.0,
|
| 206 |
+
"eval_acc_macro": 0.8256823767025346,
|
| 207 |
+
"eval_acc_micro": 0.8327024981074312,
|
| 208 |
+
"eval_auc_macro": 0.9930553828317482,
|
| 209 |
+
"eval_auc_micro": 0.9940669351858088,
|
| 210 |
+
"eval_f1_at_5": 0.4291987915672543,
|
| 211 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 212 |
+
"eval_f1_macro": 0.9023354543659394,
|
| 213 |
+
"eval_f1_micro": 0.9087154068565958,
|
| 214 |
+
"eval_loss": 0.06074792891740799,
|
| 215 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 216 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 217 |
+
"eval_prec_macro": 0.8963122585533565,
|
| 218 |
+
"eval_prec_micro": 0.9053497942386086,
|
| 219 |
+
"eval_rec_at_5": 0.9785007610350076,
|
| 220 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 221 |
+
"eval_rec_macro": 0.9088224265936838,
|
| 222 |
+
"eval_rec_micro": 0.9121061359866574,
|
| 223 |
+
"eval_runtime": 6.9556,
|
| 224 |
+
"eval_samples_per_second": 125.942,
|
| 225 |
+
"eval_steps_per_second": 15.815,
|
| 226 |
+
"step": 168
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 8.0,
|
| 230 |
+
"grad_norm": 1.107313632965088,
|
| 231 |
+
"learning_rate": 8.888888888888888e-06,
|
| 232 |
+
"loss": 0.0624,
|
| 233 |
+
"step": 192
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 8.0,
|
| 237 |
+
"eval_acc_macro": 0.8278930665915176,
|
| 238 |
+
"eval_acc_micro": 0.832705350414406,
|
| 239 |
+
"eval_auc_macro": 0.9935760753995465,
|
| 240 |
+
"eval_auc_micro": 0.9946024409081464,
|
| 241 |
+
"eval_f1_at_5": 0.4291987915672543,
|
| 242 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 243 |
+
"eval_f1_macro": 0.9037268724978929,
|
| 244 |
+
"eval_f1_micro": 0.9087171052630831,
|
| 245 |
+
"eval_loss": 0.060165293514728546,
|
| 246 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 247 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 248 |
+
"eval_prec_macro": 0.8946308202984045,
|
| 249 |
+
"eval_prec_micro": 0.9013050570961744,
|
| 250 |
+
"eval_rec_at_5": 0.9785007610350076,
|
| 251 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 252 |
+
"eval_rec_macro": 0.9137053638007582,
|
| 253 |
+
"eval_rec_micro": 0.9162520729684148,
|
| 254 |
+
"eval_runtime": 6.9888,
|
| 255 |
+
"eval_samples_per_second": 125.344,
|
| 256 |
+
"eval_steps_per_second": 15.74,
|
| 257 |
+
"step": 192
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 9.0,
|
| 261 |
+
"grad_norm": 0.9251410961151123,
|
| 262 |
+
"learning_rate": 8.611111111111112e-06,
|
| 263 |
+
"loss": 0.058,
|
| 264 |
+
"step": 216
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 9.0,
|
| 268 |
+
"eval_acc_macro": 0.8361146010604618,
|
| 269 |
+
"eval_acc_micro": 0.8392307692307046,
|
| 270 |
+
"eval_auc_macro": 0.9940701262860925,
|
| 271 |
+
"eval_auc_micro": 0.9949116264083977,
|
| 272 |
+
"eval_f1_at_5": 0.4295136788854873,
|
| 273 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 274 |
+
"eval_f1_macro": 0.9089752463509541,
|
| 275 |
+
"eval_f1_micro": 0.9125888749476442,
|
| 276 |
+
"eval_loss": 0.05600544437766075,
|
| 277 |
+
"eval_prec_at_5": 0.2751141552511416,
|
| 278 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 279 |
+
"eval_prec_macro": 0.9171011361277855,
|
| 280 |
+
"eval_prec_micro": 0.9206751054851543,
|
| 281 |
+
"eval_rec_at_5": 0.9788812785388128,
|
| 282 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 283 |
+
"eval_rec_macro": 0.90156843038978,
|
| 284 |
+
"eval_rec_micro": 0.9046434494194938,
|
| 285 |
+
"eval_runtime": 7.0672,
|
| 286 |
+
"eval_samples_per_second": 123.953,
|
| 287 |
+
"eval_steps_per_second": 15.565,
|
| 288 |
+
"step": 216
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 10.0,
|
| 292 |
+
"grad_norm": 1.4386526346206665,
|
| 293 |
+
"learning_rate": 8.333333333333334e-06,
|
| 294 |
+
"loss": 0.0556,
|
| 295 |
+
"step": 240
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 10.0,
|
| 299 |
+
"eval_acc_macro": 0.8346145293076771,
|
| 300 |
+
"eval_acc_micro": 0.8394103956554818,
|
| 301 |
+
"eval_auc_macro": 0.9940972007833064,
|
| 302 |
+
"eval_auc_micro": 0.9950393993211819,
|
| 303 |
+
"eval_f1_at_5": 0.42918048633589306,
|
| 304 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 305 |
+
"eval_f1_macro": 0.9076071562375437,
|
| 306 |
+
"eval_f1_micro": 0.9126950653731831,
|
| 307 |
+
"eval_loss": 0.05595849081873894,
|
| 308 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 309 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 310 |
+
"eval_prec_macro": 0.9241533819936333,
|
| 311 |
+
"eval_prec_micro": 0.9287553648067872,
|
| 312 |
+
"eval_rec_at_5": 0.978310502283105,
|
| 313 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 314 |
+
"eval_rec_macro": 0.8925287390946574,
|
| 315 |
+
"eval_rec_micro": 0.8971807628523302,
|
| 316 |
+
"eval_runtime": 6.9532,
|
| 317 |
+
"eval_samples_per_second": 125.984,
|
| 318 |
+
"eval_steps_per_second": 15.82,
|
| 319 |
+
"step": 240
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 11.0,
|
| 323 |
+
"grad_norm": 1.3220967054367065,
|
| 324 |
+
"learning_rate": 8.055555555555557e-06,
|
| 325 |
+
"loss": 0.0519,
|
| 326 |
+
"step": 264
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 11.0,
|
| 330 |
+
"eval_acc_macro": 0.8442462130549724,
|
| 331 |
+
"eval_acc_micro": 0.8494957331263887,
|
| 332 |
+
"eval_auc_macro": 0.9943053813562361,
|
| 333 |
+
"eval_auc_micro": 0.995346713669991,
|
| 334 |
+
"eval_f1_at_5": 0.4288655829720509,
|
| 335 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 336 |
+
"eval_f1_macro": 0.9133293255310709,
|
| 337 |
+
"eval_f1_micro": 0.9186241610737484,
|
| 338 |
+
"eval_loss": 0.05388868972659111,
|
| 339 |
+
"eval_prec_at_5": 0.27465753424657535,
|
| 340 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 341 |
+
"eval_prec_macro": 0.9247426584222537,
|
| 342 |
+
"eval_prec_micro": 0.9295415959252181,
|
| 343 |
+
"eval_rec_at_5": 0.9779299847792998,
|
| 344 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 345 |
+
"eval_rec_macro": 0.9033312145477272,
|
| 346 |
+
"eval_rec_micro": 0.9079601990048998,
|
| 347 |
+
"eval_runtime": 6.9787,
|
| 348 |
+
"eval_samples_per_second": 125.524,
|
| 349 |
+
"eval_steps_per_second": 15.762,
|
| 350 |
+
"step": 264
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 12.0,
|
| 354 |
+
"grad_norm": 0.9175763130187988,
|
| 355 |
+
"learning_rate": 7.77777777777778e-06,
|
| 356 |
+
"loss": 0.0492,
|
| 357 |
+
"step": 288
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 12.0,
|
| 361 |
+
"eval_acc_macro": 0.8470976768477376,
|
| 362 |
+
"eval_acc_micro": 0.8500376789750678,
|
| 363 |
+
"eval_auc_macro": 0.994169602401962,
|
| 364 |
+
"eval_auc_micro": 0.9953780777322643,
|
| 365 |
+
"eval_f1_at_5": 0.42853236537508876,
|
| 366 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 367 |
+
"eval_f1_macro": 0.9152662917583787,
|
| 368 |
+
"eval_f1_micro": 0.9189409368634688,
|
| 369 |
+
"eval_loss": 0.056095000356435776,
|
| 370 |
+
"eval_prec_at_5": 0.2744292237442923,
|
| 371 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 372 |
+
"eval_prec_macro": 0.9004415240992042,
|
| 373 |
+
"eval_prec_micro": 0.9031224979983263,
|
| 374 |
+
"eval_rec_at_5": 0.977359208523592,
|
| 375 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 376 |
+
"eval_rec_macro": 0.9315482026514185,
|
| 377 |
+
"eval_rec_micro": 0.9353233830844995,
|
| 378 |
+
"eval_runtime": 6.9839,
|
| 379 |
+
"eval_samples_per_second": 125.432,
|
| 380 |
+
"eval_steps_per_second": 15.751,
|
| 381 |
+
"step": 288
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 13.0,
|
| 385 |
+
"grad_norm": 1.00278639793396,
|
| 386 |
+
"learning_rate": 7.500000000000001e-06,
|
| 387 |
+
"loss": 0.048,
|
| 388 |
+
"step": 312
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 13.0,
|
| 392 |
+
"eval_acc_macro": 0.8486995058416006,
|
| 393 |
+
"eval_acc_micro": 0.8536776212831882,
|
| 394 |
+
"eval_auc_macro": 0.9946152044988873,
|
| 395 |
+
"eval_auc_micro": 0.9955295518966525,
|
| 396 |
+
"eval_f1_at_5": 0.42918048633589306,
|
| 397 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 398 |
+
"eval_f1_macro": 0.9163503224179071,
|
| 399 |
+
"eval_f1_micro": 0.921063739974595,
|
| 400 |
+
"eval_loss": 0.05099855735898018,
|
| 401 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 402 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 403 |
+
"eval_prec_macro": 0.9313959829408701,
|
| 404 |
+
"eval_prec_micro": 0.9380911435940723,
|
| 405 |
+
"eval_rec_at_5": 0.978310502283105,
|
| 406 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 407 |
+
"eval_rec_macro": 0.9025756160114312,
|
| 408 |
+
"eval_rec_micro": 0.9046434494194938,
|
| 409 |
+
"eval_runtime": 6.9796,
|
| 410 |
+
"eval_samples_per_second": 125.509,
|
| 411 |
+
"eval_steps_per_second": 15.76,
|
| 412 |
+
"step": 312
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 14.0,
|
| 416 |
+
"grad_norm": 0.7232435941696167,
|
| 417 |
+
"learning_rate": 7.222222222222223e-06,
|
| 418 |
+
"loss": 0.0457,
|
| 419 |
+
"step": 336
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 14.0,
|
| 423 |
+
"eval_acc_macro": 0.8525145546996621,
|
| 424 |
+
"eval_acc_micro": 0.8559322033897645,
|
| 425 |
+
"eval_auc_macro": 0.9948837193480031,
|
| 426 |
+
"eval_auc_micro": 0.9959032475704436,
|
| 427 |
+
"eval_f1_at_5": 0.4291987915672543,
|
| 428 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 429 |
+
"eval_f1_macro": 0.9187720987706429,
|
| 430 |
+
"eval_f1_micro": 0.9223744292236677,
|
| 431 |
+
"eval_loss": 0.04985074698925018,
|
| 432 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 433 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 434 |
+
"eval_prec_macro": 0.9187488547711858,
|
| 435 |
+
"eval_prec_micro": 0.9235245220281859,
|
| 436 |
+
"eval_rec_at_5": 0.9785007610350076,
|
| 437 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 438 |
+
"eval_rec_macro": 0.9190659938481747,
|
| 439 |
+
"eval_rec_micro": 0.9212271973465239,
|
| 440 |
+
"eval_runtime": 6.9629,
|
| 441 |
+
"eval_samples_per_second": 125.809,
|
| 442 |
+
"eval_steps_per_second": 15.798,
|
| 443 |
+
"step": 336
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 15.0,
|
| 447 |
+
"grad_norm": 0.7620428204536438,
|
| 448 |
+
"learning_rate": 6.944444444444445e-06,
|
| 449 |
+
"loss": 0.0444,
|
| 450 |
+
"step": 360
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 15.0,
|
| 454 |
+
"eval_acc_macro": 0.8563973908810851,
|
| 455 |
+
"eval_acc_micro": 0.8611532625189028,
|
| 456 |
+
"eval_auc_macro": 0.9949728097109144,
|
| 457 |
+
"eval_auc_micro": 0.9959143853766486,
|
| 458 |
+
"eval_f1_at_5": 0.42918048633589306,
|
| 459 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 460 |
+
"eval_f1_macro": 0.9209429321639129,
|
| 461 |
+
"eval_f1_micro": 0.9253974724825988,
|
| 462 |
+
"eval_loss": 0.051042765378952026,
|
| 463 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 464 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 465 |
+
"eval_prec_macro": 0.9035271919581529,
|
| 466 |
+
"eval_prec_micro": 0.9101844426623167,
|
| 467 |
+
"eval_rec_at_5": 0.978310502283105,
|
| 468 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 469 |
+
"eval_rec_macro": 0.9396785113976804,
|
| 470 |
+
"eval_rec_micro": 0.9411276948589601,
|
| 471 |
+
"eval_runtime": 7.0373,
|
| 472 |
+
"eval_samples_per_second": 124.48,
|
| 473 |
+
"eval_steps_per_second": 15.631,
|
| 474 |
+
"step": 360
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 16.0,
|
| 478 |
+
"grad_norm": 1.0121941566467285,
|
| 479 |
+
"learning_rate": 6.666666666666667e-06,
|
| 480 |
+
"loss": 0.042,
|
| 481 |
+
"step": 384
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 16.0,
|
| 485 |
+
"eval_acc_macro": 0.8578953265912642,
|
| 486 |
+
"eval_acc_micro": 0.8632148377124526,
|
| 487 |
+
"eval_auc_macro": 0.9950318833304016,
|
| 488 |
+
"eval_auc_micro": 0.9960295057415836,
|
| 489 |
+
"eval_f1_at_5": 0.42918048633589306,
|
| 490 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 491 |
+
"eval_f1_macro": 0.9219762526743717,
|
| 492 |
+
"eval_f1_micro": 0.9265864786394916,
|
| 493 |
+
"eval_loss": 0.04983897507190704,
|
| 494 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 495 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 496 |
+
"eval_prec_macro": 0.9224728362696459,
|
| 497 |
+
"eval_prec_micro": 0.9269709543567695,
|
| 498 |
+
"eval_rec_at_5": 0.978310502283105,
|
| 499 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 500 |
+
"eval_rec_macro": 0.9228006946898867,
|
| 501 |
+
"eval_rec_micro": 0.9262023217246329,
|
| 502 |
+
"eval_runtime": 6.9735,
|
| 503 |
+
"eval_samples_per_second": 125.619,
|
| 504 |
+
"eval_steps_per_second": 15.774,
|
| 505 |
+
"step": 384
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 17.0,
|
| 509 |
+
"grad_norm": 0.6501286029815674,
|
| 510 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 511 |
+
"loss": 0.0401,
|
| 512 |
+
"step": 408
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 17.0,
|
| 516 |
+
"eval_acc_macro": 0.8609431757906035,
|
| 517 |
+
"eval_acc_micro": 0.8669760247485795,
|
| 518 |
+
"eval_auc_macro": 0.9951994943771586,
|
| 519 |
+
"eval_auc_micro": 0.996281264713042,
|
| 520 |
+
"eval_f1_at_5": 0.42918048633589306,
|
| 521 |
+
"eval_f1_at_8": 0.2925113305987028,
|
| 522 |
+
"eval_f1_macro": 0.9235158618894997,
|
| 523 |
+
"eval_f1_micro": 0.9287489643744051,
|
| 524 |
+
"eval_loss": 0.04911518469452858,
|
| 525 |
+
"eval_prec_at_5": 0.27488584474885847,
|
| 526 |
+
"eval_prec_at_8": 0.17194634703196346,
|
| 527 |
+
"eval_prec_macro": 0.9229370933414865,
|
| 528 |
+
"eval_prec_micro": 0.9279801324502542,
|
| 529 |
+
"eval_rec_at_5": 0.978310502283105,
|
| 530 |
+
"eval_rec_at_8": 0.9788812785388128,
|
| 531 |
+
"eval_rec_macro": 0.9243968142875807,
|
| 532 |
+
"eval_rec_micro": 0.929519071310039,
|
| 533 |
+
"eval_runtime": 6.9496,
|
| 534 |
+
"eval_samples_per_second": 126.05,
|
| 535 |
+
"eval_steps_per_second": 15.828,
|
| 536 |
+
"step": 408
|
| 537 |
+
}
|
| 538 |
+
],
|
| 539 |
+
"logging_steps": 500,
|
| 540 |
+
"max_steps": 960,
|
| 541 |
+
"num_input_tokens_seen": 0,
|
| 542 |
+
"num_train_epochs": 40,
|
| 543 |
+
"save_steps": 500,
|
| 544 |
+
"total_flos": 1.3730832326295552e+16,
|
| 545 |
+
"train_batch_size": 32,
|
| 546 |
+
"trial_name": null,
|
| 547 |
+
"trial_params": null
|
| 548 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05f3b4d5b72f9b42cb3e2e844517f10edb36635dcf5d72af35f8401a9ceefd80
|
| 3 |
+
size 5048
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|