Add files using upload-large-folder tool
Browse files- added_tokens.json +6 -0
- checkpoint-100000/added_tokens.json +6 -0
- checkpoint-100000/gliner_config.json +137 -0
- checkpoint-100000/optimizer.pt +3 -0
- checkpoint-100000/pytorch_model.bin +3 -0
- checkpoint-100000/rng_state.pth +3 -0
- checkpoint-100000/scheduler.pt +3 -0
- checkpoint-100000/special_tokens_map.json +15 -0
- checkpoint-100000/spm.model +3 -0
- checkpoint-100000/tokenizer.json +0 -0
- checkpoint-100000/tokenizer_config.json +82 -0
- checkpoint-100000/trainer_state.json +728 -0
- checkpoint-50000/added_tokens.json +6 -0
- checkpoint-50000/gliner_config.json +137 -0
- checkpoint-50000/optimizer.pt +3 -0
- checkpoint-50000/pytorch_model.bin +3 -0
- checkpoint-50000/rng_state.pth +3 -0
- checkpoint-50000/scheduler.pt +3 -0
- checkpoint-50000/special_tokens_map.json +15 -0
- checkpoint-50000/spm.model +3 -0
- checkpoint-50000/tokenizer.json +0 -0
- checkpoint-50000/tokenizer_config.json +82 -0
- checkpoint-50000/trainer_state.json +378 -0
- gliner_config.json +137 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +15 -0
- spm.model +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +82 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<<ENT>>": 128002,
|
| 3 |
+
"<<SEP>>": 128003,
|
| 4 |
+
"[FLERT]": 128001,
|
| 5 |
+
"[MASK]": 128000
|
| 6 |
+
}
|
checkpoint-100000/added_tokens.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<<ENT>>": 128002,
|
| 3 |
+
"<<SEP>>": 128003,
|
| 4 |
+
"[FLERT]": 128001,
|
| 5 |
+
"[MASK]": 128000
|
| 6 |
+
}
|
checkpoint-100000/gliner_config.json
ADDED
|
@@ -0,0 +1,137 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"class_token_index": 128002,
|
| 3 |
+
"dropout": 0.1,
|
| 4 |
+
"embed_ent_token": true,
|
| 5 |
+
"encoder_config": {
|
| 6 |
+
"_name_or_path": "microsoft/deberta-v3-large",
|
| 7 |
+
"add_cross_attention": false,
|
| 8 |
+
"architectures": null,
|
| 9 |
+
"attention_probs_dropout_prob": 0.1,
|
| 10 |
+
"bad_words_ids": null,
|
| 11 |
+
"begin_suppress_tokens": null,
|
| 12 |
+
"bos_token_id": null,
|
| 13 |
+
"chunk_size_feed_forward": 0,
|
| 14 |
+
"cross_attention_hidden_size": null,
|
| 15 |
+
"decoder_start_token_id": null,
|
| 16 |
+
"diversity_penalty": 0.0,
|
| 17 |
+
"do_sample": false,
|
| 18 |
+
"early_stopping": false,
|
| 19 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 20 |
+
"eos_token_id": null,
|
| 21 |
+
"exponential_decay_length_penalty": null,
|
| 22 |
+
"finetuning_task": null,
|
| 23 |
+
"forced_bos_token_id": null,
|
| 24 |
+
"forced_eos_token_id": null,
|
| 25 |
+
"hidden_act": "gelu",
|
| 26 |
+
"hidden_dropout_prob": 0.1,
|
| 27 |
+
"hidden_size": 1024,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1"
|
| 31 |
+
},
|
| 32 |
+
"initializer_range": 0.02,
|
| 33 |
+
"intermediate_size": 4096,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": false,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1
|
| 39 |
+
},
|
| 40 |
+
"layer_norm_eps": 1e-07,
|
| 41 |
+
"length_penalty": 1.0,
|
| 42 |
+
"max_length": 20,
|
| 43 |
+
"max_position_embeddings": 512,
|
| 44 |
+
"max_relative_positions": -1,
|
| 45 |
+
"min_length": 0,
|
| 46 |
+
"model_type": "deberta-v2",
|
| 47 |
+
"no_repeat_ngram_size": 0,
|
| 48 |
+
"norm_rel_ebd": "layer_norm",
|
| 49 |
+
"num_attention_heads": 16,
|
| 50 |
+
"num_beam_groups": 1,
|
| 51 |
+
"num_beams": 1,
|
| 52 |
+
"num_hidden_layers": 24,
|
| 53 |
+
"num_return_sequences": 1,
|
| 54 |
+
"output_attentions": false,
|
| 55 |
+
"output_hidden_states": false,
|
| 56 |
+
"output_scores": false,
|
| 57 |
+
"pad_token_id": 0,
|
| 58 |
+
"pooler_dropout": 0,
|
| 59 |
+
"pooler_hidden_act": "gelu",
|
| 60 |
+
"pooler_hidden_size": 1024,
|
| 61 |
+
"pos_att_type": [
|
| 62 |
+
"p2c",
|
| 63 |
+
"c2p"
|
| 64 |
+
],
|
| 65 |
+
"position_biased_input": false,
|
| 66 |
+
"position_buckets": 256,
|
| 67 |
+
"prefix": null,
|
| 68 |
+
"problem_type": null,
|
| 69 |
+
"pruned_heads": {},
|
| 70 |
+
"relative_attention": true,
|
| 71 |
+
"remove_invalid_values": false,
|
| 72 |
+
"repetition_penalty": 1.0,
|
| 73 |
+
"return_dict": true,
|
| 74 |
+
"return_dict_in_generate": false,
|
| 75 |
+
"sep_token_id": null,
|
| 76 |
+
"share_att_key": true,
|
| 77 |
+
"suppress_tokens": null,
|
| 78 |
+
"task_specific_params": null,
|
| 79 |
+
"temperature": 1.0,
|
| 80 |
+
"tf_legacy_loss": false,
|
| 81 |
+
"tie_encoder_decoder": false,
|
| 82 |
+
"tie_word_embeddings": true,
|
| 83 |
+
"tokenizer_class": null,
|
| 84 |
+
"top_k": 50,
|
| 85 |
+
"top_p": 1.0,
|
| 86 |
+
"torch_dtype": null,
|
| 87 |
+
"torchscript": false,
|
| 88 |
+
"type_vocab_size": 0,
|
| 89 |
+
"typical_p": 1.0,
|
| 90 |
+
"use_bfloat16": false,
|
| 91 |
+
"vocab_size": 128004
|
| 92 |
+
},
|
| 93 |
+
"ent_token": "<<ENT>>",
|
| 94 |
+
"eval_every": 1000,
|
| 95 |
+
"fine_tune": true,
|
| 96 |
+
"freeze_token_rep": false,
|
| 97 |
+
"fuse_layers": false,
|
| 98 |
+
"has_rnn": true,
|
| 99 |
+
"hidden_size": 512,
|
| 100 |
+
"labels_encoder": null,
|
| 101 |
+
"labels_encoder_config": null,
|
| 102 |
+
"log_dir": "logs_final",
|
| 103 |
+
"loss_alpha": 0.75,
|
| 104 |
+
"loss_gamma": 0,
|
| 105 |
+
"loss_reduction": "sum",
|
| 106 |
+
"lr_encoder": "5e-6",
|
| 107 |
+
"lr_others": "7e-6",
|
| 108 |
+
"max_len": 768,
|
| 109 |
+
"max_neg_type_ratio": 1,
|
| 110 |
+
"max_types": 30,
|
| 111 |
+
"max_width": 100,
|
| 112 |
+
"model_name": "microsoft/deberta-v3-large",
|
| 113 |
+
"model_type": "gliner",
|
| 114 |
+
"name": "token level gliner large",
|
| 115 |
+
"num_post_fusion_layers": 1,
|
| 116 |
+
"num_steps": 6000,
|
| 117 |
+
"post_fusion_schema": "",
|
| 118 |
+
"prev_path": "logs_large/model_120000",
|
| 119 |
+
"random_drop": true,
|
| 120 |
+
"root_dir": "gliner_logs",
|
| 121 |
+
"save_total_limit": 10,
|
| 122 |
+
"scheduler_type": "linear",
|
| 123 |
+
"sep_token": "<<SEP>>",
|
| 124 |
+
"shuffle_types": true,
|
| 125 |
+
"size_sup": -1,
|
| 126 |
+
"span_mode": "token_level",
|
| 127 |
+
"subtoken_pooling": "first",
|
| 128 |
+
"train_batch_size": 8,
|
| 129 |
+
"train_data": "../../data/unie_ner1.json",
|
| 130 |
+
"transformers_version": "4.40.1",
|
| 131 |
+
"val_data_dir": "none",
|
| 132 |
+
"vocab_size": 128004,
|
| 133 |
+
"warmup_ratio": 0.1,
|
| 134 |
+
"weight_decay_encoder": 0.01,
|
| 135 |
+
"weight_decay_other": 0.01,
|
| 136 |
+
"words_splitter_type": "whitespace"
|
| 137 |
+
}
|
checkpoint-100000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a589be94e7f5278130f05f09171a1cb01e5fbf42ee958d9e5220a88ef8eee907
|
| 3 |
+
size 3522103130
|
checkpoint-100000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e5dd8f5513c8424eb5dd3386f2af3becc0174cdbe31aa87db1ac392ec7f0d7b
|
| 3 |
+
size 1761030542
|
checkpoint-100000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a484153e33374131f1f0e9fbeaa8a970d0fc217e52ff875cabbd203b10169790
|
| 3 |
+
size 14244
|
checkpoint-100000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1e4f9440be972605160ad13669492587524fba31895fe9ff29b61da0f9d0bba
|
| 3 |
+
size 1064
|
checkpoint-100000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"eos_token": "[SEP]",
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"pad_token": "[PAD]",
|
| 7 |
+
"sep_token": "[SEP]",
|
| 8 |
+
"unk_token": {
|
| 9 |
+
"content": "[UNK]",
|
| 10 |
+
"lstrip": false,
|
| 11 |
+
"normalized": true,
|
| 12 |
+
"rstrip": false,
|
| 13 |
+
"single_word": false
|
| 14 |
+
}
|
| 15 |
+
}
|
checkpoint-100000/spm.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
| 3 |
+
size 2464616
|
checkpoint-100000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-100000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "[CLS]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[SEP]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[UNK]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"128000": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"128001": {
|
| 44 |
+
"content": "[FLERT]",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
"128002": {
|
| 52 |
+
"content": "<<ENT>>",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
+
},
|
| 59 |
+
"128003": {
|
| 60 |
+
"content": "<<SEP>>",
|
| 61 |
+
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
+
}
|
| 67 |
+
},
|
| 68 |
+
"bos_token": "[CLS]",
|
| 69 |
+
"clean_up_tokenization_spaces": true,
|
| 70 |
+
"cls_token": "[CLS]",
|
| 71 |
+
"do_lower_case": false,
|
| 72 |
+
"eos_token": "[SEP]",
|
| 73 |
+
"mask_token": "[MASK]",
|
| 74 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 75 |
+
"pad_token": "[PAD]",
|
| 76 |
+
"sep_token": "[SEP]",
|
| 77 |
+
"sp_model_kwargs": {},
|
| 78 |
+
"split_by_punct": false,
|
| 79 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 80 |
+
"unk_token": "[UNK]",
|
| 81 |
+
"vocab_type": "spm"
|
| 82 |
+
}
|
checkpoint-100000/trainer_state.json
ADDED
|
@@ -0,0 +1,728 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.7314914378927194,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 100000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 7.314914378927195e-06,
|
| 13 |
+
"grad_norm": 2702.876220703125,
|
| 14 |
+
"learning_rate": 3.6573769292663306e-10,
|
| 15 |
+
"loss": 185.8854,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.007314914378927195,
|
| 20 |
+
"grad_norm": 25910.705078125,
|
| 21 |
+
"learning_rate": 3.6573769292663306e-07,
|
| 22 |
+
"loss": 48.434,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01462982875785439,
|
| 27 |
+
"grad_norm": 3.9086620807647705,
|
| 28 |
+
"learning_rate": 7.314753858532661e-07,
|
| 29 |
+
"loss": 17.9804,
|
| 30 |
+
"step": 2000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.021944743136781585,
|
| 34 |
+
"grad_norm": 17.556304931640625,
|
| 35 |
+
"learning_rate": 1.097213078779899e-06,
|
| 36 |
+
"loss": 10.5262,
|
| 37 |
+
"step": 3000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02925965751570878,
|
| 41 |
+
"grad_norm": 4689.71240234375,
|
| 42 |
+
"learning_rate": 1.4629507717065323e-06,
|
| 43 |
+
"loss": 11.6651,
|
| 44 |
+
"step": 4000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03657457189463598,
|
| 48 |
+
"grad_norm": 0.0065016308799386024,
|
| 49 |
+
"learning_rate": 1.8286884646331652e-06,
|
| 50 |
+
"loss": 9.4746,
|
| 51 |
+
"step": 5000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04388948627356317,
|
| 55 |
+
"grad_norm": 0.060599055141210556,
|
| 56 |
+
"learning_rate": 2.194426157559798e-06,
|
| 57 |
+
"loss": 7.1303,
|
| 58 |
+
"step": 6000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.051204400652490364,
|
| 62 |
+
"grad_norm": 1028.9453125,
|
| 63 |
+
"learning_rate": 2.560163850486431e-06,
|
| 64 |
+
"loss": 8.917,
|
| 65 |
+
"step": 7000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05851931503141756,
|
| 69 |
+
"grad_norm": 5.463726043701172,
|
| 70 |
+
"learning_rate": 2.9259015434130645e-06,
|
| 71 |
+
"loss": 8.9339,
|
| 72 |
+
"step": 8000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06583422941034475,
|
| 76 |
+
"grad_norm": 0.3307730257511139,
|
| 77 |
+
"learning_rate": 3.2916392363396975e-06,
|
| 78 |
+
"loss": 8.746,
|
| 79 |
+
"step": 9000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07314914378927195,
|
| 83 |
+
"grad_norm": 0.003347629914060235,
|
| 84 |
+
"learning_rate": 3.6573769292663304e-06,
|
| 85 |
+
"loss": 9.4711,
|
| 86 |
+
"step": 10000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08046405816819914,
|
| 90 |
+
"grad_norm": 0.006155295763164759,
|
| 91 |
+
"learning_rate": 4.023114622192964e-06,
|
| 92 |
+
"loss": 7.647,
|
| 93 |
+
"step": 11000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08777897254712634,
|
| 97 |
+
"grad_norm": 0.0019895241130143404,
|
| 98 |
+
"learning_rate": 4.388852315119596e-06,
|
| 99 |
+
"loss": 5.7327,
|
| 100 |
+
"step": 12000
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.09509388692605353,
|
| 104 |
+
"grad_norm": 0.004527593031525612,
|
| 105 |
+
"learning_rate": 4.75459000804623e-06,
|
| 106 |
+
"loss": 5.5751,
|
| 107 |
+
"step": 13000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.10240880130498073,
|
| 111 |
+
"grad_norm": 0.009128883481025696,
|
| 112 |
+
"learning_rate": 4.986629929451543e-06,
|
| 113 |
+
"loss": 8.6148,
|
| 114 |
+
"step": 14000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.10972371568390792,
|
| 118 |
+
"grad_norm": 0.012683026492595673,
|
| 119 |
+
"learning_rate": 4.9459914171462015e-06,
|
| 120 |
+
"loss": 8.5558,
|
| 121 |
+
"step": 15000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11703863006283512,
|
| 125 |
+
"grad_norm": 0.0028167981654405594,
|
| 126 |
+
"learning_rate": 4.90535290484086e-06,
|
| 127 |
+
"loss": 6.0433,
|
| 128 |
+
"step": 16000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1243535444417623,
|
| 132 |
+
"grad_norm": 0.0011900264071300626,
|
| 133 |
+
"learning_rate": 4.864714392535519e-06,
|
| 134 |
+
"loss": 6.9084,
|
| 135 |
+
"step": 17000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1316684588206895,
|
| 139 |
+
"grad_norm": 0.025524910539388657,
|
| 140 |
+
"learning_rate": 4.824075880230177e-06,
|
| 141 |
+
"loss": 6.7333,
|
| 142 |
+
"step": 18000
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1389833731996167,
|
| 146 |
+
"grad_norm": 0.027619725093245506,
|
| 147 |
+
"learning_rate": 4.783437367924835e-06,
|
| 148 |
+
"loss": 4.1436,
|
| 149 |
+
"step": 19000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1462982875785439,
|
| 153 |
+
"grad_norm": 0.012698939070105553,
|
| 154 |
+
"learning_rate": 4.742798855619494e-06,
|
| 155 |
+
"loss": 5.397,
|
| 156 |
+
"step": 20000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.15361320195747108,
|
| 160 |
+
"grad_norm": 0.031142177060246468,
|
| 161 |
+
"learning_rate": 4.702160343314152e-06,
|
| 162 |
+
"loss": 5.5156,
|
| 163 |
+
"step": 21000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.16092811633639828,
|
| 167 |
+
"grad_norm": 0.0004189134924672544,
|
| 168 |
+
"learning_rate": 4.66152183100881e-06,
|
| 169 |
+
"loss": 4.8633,
|
| 170 |
+
"step": 22000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.16824303071532548,
|
| 174 |
+
"grad_norm": 0.0103899035602808,
|
| 175 |
+
"learning_rate": 4.620883318703469e-06,
|
| 176 |
+
"loss": 7.2146,
|
| 177 |
+
"step": 23000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17555794509425268,
|
| 181 |
+
"grad_norm": 1158.8017578125,
|
| 182 |
+
"learning_rate": 4.580244806398127e-06,
|
| 183 |
+
"loss": 5.7667,
|
| 184 |
+
"step": 24000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18287285947317986,
|
| 188 |
+
"grad_norm": 0.03399639576673508,
|
| 189 |
+
"learning_rate": 4.5396062940927856e-06,
|
| 190 |
+
"loss": 5.0472,
|
| 191 |
+
"step": 25000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.19018777385210706,
|
| 195 |
+
"grad_norm": 0.017644532024860382,
|
| 196 |
+
"learning_rate": 4.4989677817874446e-06,
|
| 197 |
+
"loss": 4.8188,
|
| 198 |
+
"step": 26000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.19750268823103426,
|
| 202 |
+
"grad_norm": 0.00010079160711029544,
|
| 203 |
+
"learning_rate": 4.4583292694821035e-06,
|
| 204 |
+
"loss": 5.7598,
|
| 205 |
+
"step": 27000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.20481760260996146,
|
| 209 |
+
"grad_norm": 0.009618501178920269,
|
| 210 |
+
"learning_rate": 4.417690757176762e-06,
|
| 211 |
+
"loss": 4.683,
|
| 212 |
+
"step": 28000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.21213251698888866,
|
| 216 |
+
"grad_norm": 0.018396975472569466,
|
| 217 |
+
"learning_rate": 4.377052244871421e-06,
|
| 218 |
+
"loss": 5.7816,
|
| 219 |
+
"step": 29000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.21944743136781583,
|
| 223 |
+
"grad_norm": 0.026549218222498894,
|
| 224 |
+
"learning_rate": 4.336413732566079e-06,
|
| 225 |
+
"loss": 5.5149,
|
| 226 |
+
"step": 30000
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.22676234574674303,
|
| 230 |
+
"grad_norm": 0.01402178592979908,
|
| 231 |
+
"learning_rate": 4.295775220260737e-06,
|
| 232 |
+
"loss": 6.1021,
|
| 233 |
+
"step": 31000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.23407726012567023,
|
| 237 |
+
"grad_norm": 2950.190185546875,
|
| 238 |
+
"learning_rate": 4.255136707955396e-06,
|
| 239 |
+
"loss": 5.1742,
|
| 240 |
+
"step": 32000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.24139217450459743,
|
| 244 |
+
"grad_norm": 0.01243713591247797,
|
| 245 |
+
"learning_rate": 4.214498195650054e-06,
|
| 246 |
+
"loss": 4.8856,
|
| 247 |
+
"step": 33000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2487070888835246,
|
| 251 |
+
"grad_norm": 0.00121857482008636,
|
| 252 |
+
"learning_rate": 4.173859683344712e-06,
|
| 253 |
+
"loss": 4.0296,
|
| 254 |
+
"step": 34000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.25602200326245184,
|
| 258 |
+
"grad_norm": 0.021528728306293488,
|
| 259 |
+
"learning_rate": 4.133221171039371e-06,
|
| 260 |
+
"loss": 3.7989,
|
| 261 |
+
"step": 35000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.263336917641379,
|
| 265 |
+
"grad_norm": 0.067794568836689,
|
| 266 |
+
"learning_rate": 4.092582658734029e-06,
|
| 267 |
+
"loss": 4.8373,
|
| 268 |
+
"step": 36000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2706518320203062,
|
| 272 |
+
"grad_norm": 83.66200256347656,
|
| 273 |
+
"learning_rate": 4.0519441464286876e-06,
|
| 274 |
+
"loss": 3.2441,
|
| 275 |
+
"step": 37000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.2779667463992334,
|
| 279 |
+
"grad_norm": 0.00217541866004467,
|
| 280 |
+
"learning_rate": 4.0113056341233466e-06,
|
| 281 |
+
"loss": 3.5578,
|
| 282 |
+
"step": 38000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2852816607781606,
|
| 286 |
+
"grad_norm": 0.0008728219545446336,
|
| 287 |
+
"learning_rate": 3.970667121818005e-06,
|
| 288 |
+
"loss": 2.6644,
|
| 289 |
+
"step": 39000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2925965751570878,
|
| 293 |
+
"grad_norm": 0.027021408081054688,
|
| 294 |
+
"learning_rate": 3.930028609512664e-06,
|
| 295 |
+
"loss": 3.7778,
|
| 296 |
+
"step": 40000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.299911489536015,
|
| 300 |
+
"grad_norm": 0.001312136766500771,
|
| 301 |
+
"learning_rate": 3.889390097207322e-06,
|
| 302 |
+
"loss": 4.2509,
|
| 303 |
+
"step": 41000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.30722640391494216,
|
| 307 |
+
"grad_norm": 0.011145360767841339,
|
| 308 |
+
"learning_rate": 3.84875158490198e-06,
|
| 309 |
+
"loss": 4.6084,
|
| 310 |
+
"step": 42000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3145413182938694,
|
| 314 |
+
"grad_norm": 289.0566711425781,
|
| 315 |
+
"learning_rate": 3.8081130725966386e-06,
|
| 316 |
+
"loss": 3.3176,
|
| 317 |
+
"step": 43000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.32185623267279656,
|
| 321 |
+
"grad_norm": 0.006871068850159645,
|
| 322 |
+
"learning_rate": 3.767474560291297e-06,
|
| 323 |
+
"loss": 3.0108,
|
| 324 |
+
"step": 44000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.32917114705172373,
|
| 328 |
+
"grad_norm": 0.002771923318505287,
|
| 329 |
+
"learning_rate": 3.7268360479859557e-06,
|
| 330 |
+
"loss": 3.5366,
|
| 331 |
+
"step": 45000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.33648606143065096,
|
| 335 |
+
"grad_norm": 0.0030192858539521694,
|
| 336 |
+
"learning_rate": 3.686197535680614e-06,
|
| 337 |
+
"loss": 3.3997,
|
| 338 |
+
"step": 46000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.34380097580957814,
|
| 342 |
+
"grad_norm": 3926.616943359375,
|
| 343 |
+
"learning_rate": 3.6455590233752724e-06,
|
| 344 |
+
"loss": 3.5042,
|
| 345 |
+
"step": 47000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.35111589018850536,
|
| 349 |
+
"grad_norm": 0.000544128124602139,
|
| 350 |
+
"learning_rate": 3.604920511069931e-06,
|
| 351 |
+
"loss": 2.9953,
|
| 352 |
+
"step": 48000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.35843080456743254,
|
| 356 |
+
"grad_norm": 14144.9677734375,
|
| 357 |
+
"learning_rate": 3.564281998764589e-06,
|
| 358 |
+
"loss": 2.9742,
|
| 359 |
+
"step": 49000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.3657457189463597,
|
| 363 |
+
"grad_norm": 0.0021343908738344908,
|
| 364 |
+
"learning_rate": 3.5236434864592477e-06,
|
| 365 |
+
"loss": 3.4145,
|
| 366 |
+
"step": 50000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.37306063332528694,
|
| 370 |
+
"grad_norm": 0.021554578095674515,
|
| 371 |
+
"learning_rate": 3.4830049741539063e-06,
|
| 372 |
+
"loss": 3.8241,
|
| 373 |
+
"step": 51000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.3803755477042141,
|
| 377 |
+
"grad_norm": 0.020491141825914383,
|
| 378 |
+
"learning_rate": 3.442366461848565e-06,
|
| 379 |
+
"loss": 3.6484,
|
| 380 |
+
"step": 52000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.38769046208314134,
|
| 384 |
+
"grad_norm": 0.00027259966009296477,
|
| 385 |
+
"learning_rate": 3.401727949543223e-06,
|
| 386 |
+
"loss": 3.2795,
|
| 387 |
+
"step": 53000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.3950053764620685,
|
| 391 |
+
"grad_norm": 537.7772827148438,
|
| 392 |
+
"learning_rate": 3.3610894372378816e-06,
|
| 393 |
+
"loss": 3.4838,
|
| 394 |
+
"step": 54000
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.4023202908409957,
|
| 398 |
+
"grad_norm": 0.021593140438199043,
|
| 399 |
+
"learning_rate": 3.32045092493254e-06,
|
| 400 |
+
"loss": 2.957,
|
| 401 |
+
"step": 55000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4096352052199229,
|
| 405 |
+
"grad_norm": 0.007829766720533371,
|
| 406 |
+
"learning_rate": 3.279812412627199e-06,
|
| 407 |
+
"loss": 2.7857,
|
| 408 |
+
"step": 56000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.4169501195988501,
|
| 412 |
+
"grad_norm": 0.0030255785677582026,
|
| 413 |
+
"learning_rate": 3.2391739003218577e-06,
|
| 414 |
+
"loss": 2.7841,
|
| 415 |
+
"step": 57000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.4242650339777773,
|
| 419 |
+
"grad_norm": 0.0005273462738841772,
|
| 420 |
+
"learning_rate": 3.198535388016516e-06,
|
| 421 |
+
"loss": 3.0931,
|
| 422 |
+
"step": 58000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4315799483567045,
|
| 426 |
+
"grad_norm": 0.005731058772653341,
|
| 427 |
+
"learning_rate": 3.1578968757111744e-06,
|
| 428 |
+
"loss": 2.693,
|
| 429 |
+
"step": 59000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.43889486273563166,
|
| 433 |
+
"grad_norm": 0.07265722006559372,
|
| 434 |
+
"learning_rate": 3.117258363405833e-06,
|
| 435 |
+
"loss": 2.2466,
|
| 436 |
+
"step": 60000
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.4462097771145589,
|
| 440 |
+
"grad_norm": 0.0013541084481403232,
|
| 441 |
+
"learning_rate": 3.076619851100491e-06,
|
| 442 |
+
"loss": 2.9045,
|
| 443 |
+
"step": 61000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.45352469149348607,
|
| 447 |
+
"grad_norm": 1008.484375,
|
| 448 |
+
"learning_rate": 3.0359813387951497e-06,
|
| 449 |
+
"loss": 3.6181,
|
| 450 |
+
"step": 62000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.46083960587241324,
|
| 454 |
+
"grad_norm": 0.0007216822705231607,
|
| 455 |
+
"learning_rate": 2.9953428264898083e-06,
|
| 456 |
+
"loss": 2.8236,
|
| 457 |
+
"step": 63000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.46815452025134047,
|
| 461 |
+
"grad_norm": 0.012205916456878185,
|
| 462 |
+
"learning_rate": 2.954704314184467e-06,
|
| 463 |
+
"loss": 2.1196,
|
| 464 |
+
"step": 64000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.47546943463026764,
|
| 468 |
+
"grad_norm": 0.004514534957706928,
|
| 469 |
+
"learning_rate": 2.914065801879125e-06,
|
| 470 |
+
"loss": 3.6147,
|
| 471 |
+
"step": 65000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.48278434900919487,
|
| 475 |
+
"grad_norm": 0.04611637070775032,
|
| 476 |
+
"learning_rate": 2.8734272895737836e-06,
|
| 477 |
+
"loss": 3.051,
|
| 478 |
+
"step": 66000
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.49009926338812204,
|
| 482 |
+
"grad_norm": 0.0011849659495055676,
|
| 483 |
+
"learning_rate": 2.832788777268442e-06,
|
| 484 |
+
"loss": 3.6258,
|
| 485 |
+
"step": 67000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.4974141777670492,
|
| 489 |
+
"grad_norm": 0.0005676737637259066,
|
| 490 |
+
"learning_rate": 2.7921502649631003e-06,
|
| 491 |
+
"loss": 2.6537,
|
| 492 |
+
"step": 68000
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5047290921459764,
|
| 496 |
+
"grad_norm": 0.4254292845726013,
|
| 497 |
+
"learning_rate": 2.751511752657759e-06,
|
| 498 |
+
"loss": 3.4537,
|
| 499 |
+
"step": 69000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5120440065249037,
|
| 503 |
+
"grad_norm": 0.0008139291894622147,
|
| 504 |
+
"learning_rate": 2.7108732403524175e-06,
|
| 505 |
+
"loss": 2.5491,
|
| 506 |
+
"step": 70000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5193589209038308,
|
| 510 |
+
"grad_norm": 0.0014713549753651023,
|
| 511 |
+
"learning_rate": 2.670234728047076e-06,
|
| 512 |
+
"loss": 2.7237,
|
| 513 |
+
"step": 71000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.526673835282758,
|
| 517 |
+
"grad_norm": 0.009666742756962776,
|
| 518 |
+
"learning_rate": 2.629596215741734e-06,
|
| 519 |
+
"loss": 3.0908,
|
| 520 |
+
"step": 72000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5339887496616852,
|
| 524 |
+
"grad_norm": 0.00042533871601335704,
|
| 525 |
+
"learning_rate": 2.5889577034363927e-06,
|
| 526 |
+
"loss": 2.5412,
|
| 527 |
+
"step": 73000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5413036640406124,
|
| 531 |
+
"grad_norm": 0.01702120713889599,
|
| 532 |
+
"learning_rate": 2.5483191911310513e-06,
|
| 533 |
+
"loss": 1.9617,
|
| 534 |
+
"step": 74000
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5486185784195396,
|
| 538 |
+
"grad_norm": 0.003681492991745472,
|
| 539 |
+
"learning_rate": 2.5076806788257095e-06,
|
| 540 |
+
"loss": 2.0281,
|
| 541 |
+
"step": 75000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5559334927984668,
|
| 545 |
+
"grad_norm": 0.007156323175877333,
|
| 546 |
+
"learning_rate": 2.467042166520368e-06,
|
| 547 |
+
"loss": 2.5386,
|
| 548 |
+
"step": 76000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.563248407177394,
|
| 552 |
+
"grad_norm": 0.006489938590675592,
|
| 553 |
+
"learning_rate": 2.426403654215027e-06,
|
| 554 |
+
"loss": 1.6657,
|
| 555 |
+
"step": 77000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5705633215563212,
|
| 559 |
+
"grad_norm": 0.002306995214894414,
|
| 560 |
+
"learning_rate": 2.385765141909685e-06,
|
| 561 |
+
"loss": 3.5833,
|
| 562 |
+
"step": 78000
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5778782359352483,
|
| 566 |
+
"grad_norm": 0.009546785615384579,
|
| 567 |
+
"learning_rate": 2.3451266296043438e-06,
|
| 568 |
+
"loss": 1.7718,
|
| 569 |
+
"step": 79000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5851931503141756,
|
| 573 |
+
"grad_norm": 1.0803489203681238e-05,
|
| 574 |
+
"learning_rate": 2.3044881172990023e-06,
|
| 575 |
+
"loss": 1.7368,
|
| 576 |
+
"step": 80000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5925080646931028,
|
| 580 |
+
"grad_norm": 8.56118494994007e-05,
|
| 581 |
+
"learning_rate": 2.2638496049936605e-06,
|
| 582 |
+
"loss": 2.2237,
|
| 583 |
+
"step": 81000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.59982297907203,
|
| 587 |
+
"grad_norm": 0.001616801368072629,
|
| 588 |
+
"learning_rate": 2.223211092688319e-06,
|
| 589 |
+
"loss": 3.1469,
|
| 590 |
+
"step": 82000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6071378934509571,
|
| 594 |
+
"grad_norm": 0.049229834228754044,
|
| 595 |
+
"learning_rate": 2.1825725803829776e-06,
|
| 596 |
+
"loss": 3.9212,
|
| 597 |
+
"step": 83000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6144528078298843,
|
| 601 |
+
"grad_norm": 1679.7652587890625,
|
| 602 |
+
"learning_rate": 2.141934068077636e-06,
|
| 603 |
+
"loss": 2.5332,
|
| 604 |
+
"step": 84000
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6217677222088115,
|
| 608 |
+
"grad_norm": 0.0006012212252244353,
|
| 609 |
+
"learning_rate": 2.1012955557722943e-06,
|
| 610 |
+
"loss": 2.2774,
|
| 611 |
+
"step": 85000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.6290826365877388,
|
| 615 |
+
"grad_norm": 0.0077589754946529865,
|
| 616 |
+
"learning_rate": 2.060657043466953e-06,
|
| 617 |
+
"loss": 3.637,
|
| 618 |
+
"step": 86000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.636397550966666,
|
| 622 |
+
"grad_norm": 0.06067018210887909,
|
| 623 |
+
"learning_rate": 2.0200185311616115e-06,
|
| 624 |
+
"loss": 2.1633,
|
| 625 |
+
"step": 87000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6437124653455931,
|
| 629 |
+
"grad_norm": 0.00032135736546479166,
|
| 630 |
+
"learning_rate": 1.9793800188562696e-06,
|
| 631 |
+
"loss": 2.8641,
|
| 632 |
+
"step": 88000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6510273797245203,
|
| 636 |
+
"grad_norm": 0.00010657509119482711,
|
| 637 |
+
"learning_rate": 1.938741506550928e-06,
|
| 638 |
+
"loss": 2.6836,
|
| 639 |
+
"step": 89000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6583422941034475,
|
| 643 |
+
"grad_norm": 0.0174532663077116,
|
| 644 |
+
"learning_rate": 1.8981029942455868e-06,
|
| 645 |
+
"loss": 2.0841,
|
| 646 |
+
"step": 90000
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.6656572084823748,
|
| 650 |
+
"grad_norm": 0.00026226602494716644,
|
| 651 |
+
"learning_rate": 1.8574644819402451e-06,
|
| 652 |
+
"loss": 1.6138,
|
| 653 |
+
"step": 91000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.6729721228613019,
|
| 657 |
+
"grad_norm": 0.004689433146268129,
|
| 658 |
+
"learning_rate": 1.816825969634904e-06,
|
| 659 |
+
"loss": 1.9295,
|
| 660 |
+
"step": 92000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6802870372402291,
|
| 664 |
+
"grad_norm": 0.00042851018952205777,
|
| 665 |
+
"learning_rate": 1.7761874573295623e-06,
|
| 666 |
+
"loss": 2.4209,
|
| 667 |
+
"step": 93000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6876019516191563,
|
| 671 |
+
"grad_norm": 0.00046010102960281074,
|
| 672 |
+
"learning_rate": 1.7355489450242208e-06,
|
| 673 |
+
"loss": 2.4989,
|
| 674 |
+
"step": 94000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6949168659980834,
|
| 678 |
+
"grad_norm": 0.0018447510665282607,
|
| 679 |
+
"learning_rate": 1.6949104327188792e-06,
|
| 680 |
+
"loss": 2.1855,
|
| 681 |
+
"step": 95000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7022317803770107,
|
| 685 |
+
"grad_norm": 0.010036293417215347,
|
| 686 |
+
"learning_rate": 1.6542719204135378e-06,
|
| 687 |
+
"loss": 1.9382,
|
| 688 |
+
"step": 96000
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7095466947559379,
|
| 692 |
+
"grad_norm": 0.0004497400368563831,
|
| 693 |
+
"learning_rate": 1.6136334081081961e-06,
|
| 694 |
+
"loss": 2.08,
|
| 695 |
+
"step": 97000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.7168616091348651,
|
| 699 |
+
"grad_norm": 0.07228240370750427,
|
| 700 |
+
"learning_rate": 1.5729948958028545e-06,
|
| 701 |
+
"loss": 1.8175,
|
| 702 |
+
"step": 98000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.7241765235137922,
|
| 706 |
+
"grad_norm": 0.00016359405708499253,
|
| 707 |
+
"learning_rate": 1.532356383497513e-06,
|
| 708 |
+
"loss": 1.547,
|
| 709 |
+
"step": 99000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.7314914378927194,
|
| 713 |
+
"grad_norm": 0.010131311602890491,
|
| 714 |
+
"learning_rate": 1.4917178711921714e-06,
|
| 715 |
+
"loss": 1.5153,
|
| 716 |
+
"step": 100000
|
| 717 |
+
}
|
| 718 |
+
],
|
| 719 |
+
"logging_steps": 1000,
|
| 720 |
+
"max_steps": 136707,
|
| 721 |
+
"num_input_tokens_seen": 0,
|
| 722 |
+
"num_train_epochs": 1,
|
| 723 |
+
"save_steps": 50000,
|
| 724 |
+
"total_flos": 0.0,
|
| 725 |
+
"train_batch_size": 2,
|
| 726 |
+
"trial_name": null,
|
| 727 |
+
"trial_params": null
|
| 728 |
+
}
|
checkpoint-50000/added_tokens.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<<ENT>>": 128002,
|
| 3 |
+
"<<SEP>>": 128003,
|
| 4 |
+
"[FLERT]": 128001,
|
| 5 |
+
"[MASK]": 128000
|
| 6 |
+
}
|
checkpoint-50000/gliner_config.json
ADDED
|
@@ -0,0 +1,137 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"class_token_index": 128002,
|
| 3 |
+
"dropout": 0.1,
|
| 4 |
+
"embed_ent_token": true,
|
| 5 |
+
"encoder_config": {
|
| 6 |
+
"_name_or_path": "microsoft/deberta-v3-large",
|
| 7 |
+
"add_cross_attention": false,
|
| 8 |
+
"architectures": null,
|
| 9 |
+
"attention_probs_dropout_prob": 0.1,
|
| 10 |
+
"bad_words_ids": null,
|
| 11 |
+
"begin_suppress_tokens": null,
|
| 12 |
+
"bos_token_id": null,
|
| 13 |
+
"chunk_size_feed_forward": 0,
|
| 14 |
+
"cross_attention_hidden_size": null,
|
| 15 |
+
"decoder_start_token_id": null,
|
| 16 |
+
"diversity_penalty": 0.0,
|
| 17 |
+
"do_sample": false,
|
| 18 |
+
"early_stopping": false,
|
| 19 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 20 |
+
"eos_token_id": null,
|
| 21 |
+
"exponential_decay_length_penalty": null,
|
| 22 |
+
"finetuning_task": null,
|
| 23 |
+
"forced_bos_token_id": null,
|
| 24 |
+
"forced_eos_token_id": null,
|
| 25 |
+
"hidden_act": "gelu",
|
| 26 |
+
"hidden_dropout_prob": 0.1,
|
| 27 |
+
"hidden_size": 1024,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1"
|
| 31 |
+
},
|
| 32 |
+
"initializer_range": 0.02,
|
| 33 |
+
"intermediate_size": 4096,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": false,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1
|
| 39 |
+
},
|
| 40 |
+
"layer_norm_eps": 1e-07,
|
| 41 |
+
"length_penalty": 1.0,
|
| 42 |
+
"max_length": 20,
|
| 43 |
+
"max_position_embeddings": 512,
|
| 44 |
+
"max_relative_positions": -1,
|
| 45 |
+
"min_length": 0,
|
| 46 |
+
"model_type": "deberta-v2",
|
| 47 |
+
"no_repeat_ngram_size": 0,
|
| 48 |
+
"norm_rel_ebd": "layer_norm",
|
| 49 |
+
"num_attention_heads": 16,
|
| 50 |
+
"num_beam_groups": 1,
|
| 51 |
+
"num_beams": 1,
|
| 52 |
+
"num_hidden_layers": 24,
|
| 53 |
+
"num_return_sequences": 1,
|
| 54 |
+
"output_attentions": false,
|
| 55 |
+
"output_hidden_states": false,
|
| 56 |
+
"output_scores": false,
|
| 57 |
+
"pad_token_id": 0,
|
| 58 |
+
"pooler_dropout": 0,
|
| 59 |
+
"pooler_hidden_act": "gelu",
|
| 60 |
+
"pooler_hidden_size": 1024,
|
| 61 |
+
"pos_att_type": [
|
| 62 |
+
"p2c",
|
| 63 |
+
"c2p"
|
| 64 |
+
],
|
| 65 |
+
"position_biased_input": false,
|
| 66 |
+
"position_buckets": 256,
|
| 67 |
+
"prefix": null,
|
| 68 |
+
"problem_type": null,
|
| 69 |
+
"pruned_heads": {},
|
| 70 |
+
"relative_attention": true,
|
| 71 |
+
"remove_invalid_values": false,
|
| 72 |
+
"repetition_penalty": 1.0,
|
| 73 |
+
"return_dict": true,
|
| 74 |
+
"return_dict_in_generate": false,
|
| 75 |
+
"sep_token_id": null,
|
| 76 |
+
"share_att_key": true,
|
| 77 |
+
"suppress_tokens": null,
|
| 78 |
+
"task_specific_params": null,
|
| 79 |
+
"temperature": 1.0,
|
| 80 |
+
"tf_legacy_loss": false,
|
| 81 |
+
"tie_encoder_decoder": false,
|
| 82 |
+
"tie_word_embeddings": true,
|
| 83 |
+
"tokenizer_class": null,
|
| 84 |
+
"top_k": 50,
|
| 85 |
+
"top_p": 1.0,
|
| 86 |
+
"torch_dtype": null,
|
| 87 |
+
"torchscript": false,
|
| 88 |
+
"type_vocab_size": 0,
|
| 89 |
+
"typical_p": 1.0,
|
| 90 |
+
"use_bfloat16": false,
|
| 91 |
+
"vocab_size": 128004
|
| 92 |
+
},
|
| 93 |
+
"ent_token": "<<ENT>>",
|
| 94 |
+
"eval_every": 1000,
|
| 95 |
+
"fine_tune": true,
|
| 96 |
+
"freeze_token_rep": false,
|
| 97 |
+
"fuse_layers": false,
|
| 98 |
+
"has_rnn": true,
|
| 99 |
+
"hidden_size": 512,
|
| 100 |
+
"labels_encoder": null,
|
| 101 |
+
"labels_encoder_config": null,
|
| 102 |
+
"log_dir": "logs_final",
|
| 103 |
+
"loss_alpha": 0.75,
|
| 104 |
+
"loss_gamma": 0,
|
| 105 |
+
"loss_reduction": "sum",
|
| 106 |
+
"lr_encoder": "5e-6",
|
| 107 |
+
"lr_others": "7e-6",
|
| 108 |
+
"max_len": 768,
|
| 109 |
+
"max_neg_type_ratio": 1,
|
| 110 |
+
"max_types": 30,
|
| 111 |
+
"max_width": 100,
|
| 112 |
+
"model_name": "microsoft/deberta-v3-large",
|
| 113 |
+
"model_type": "gliner",
|
| 114 |
+
"name": "token level gliner large",
|
| 115 |
+
"num_post_fusion_layers": 1,
|
| 116 |
+
"num_steps": 6000,
|
| 117 |
+
"post_fusion_schema": "",
|
| 118 |
+
"prev_path": "logs_large/model_120000",
|
| 119 |
+
"random_drop": true,
|
| 120 |
+
"root_dir": "gliner_logs",
|
| 121 |
+
"save_total_limit": 10,
|
| 122 |
+
"scheduler_type": "linear",
|
| 123 |
+
"sep_token": "<<SEP>>",
|
| 124 |
+
"shuffle_types": true,
|
| 125 |
+
"size_sup": -1,
|
| 126 |
+
"span_mode": "token_level",
|
| 127 |
+
"subtoken_pooling": "first",
|
| 128 |
+
"train_batch_size": 8,
|
| 129 |
+
"train_data": "../../data/unie_ner1.json",
|
| 130 |
+
"transformers_version": "4.40.1",
|
| 131 |
+
"val_data_dir": "none",
|
| 132 |
+
"vocab_size": 128004,
|
| 133 |
+
"warmup_ratio": 0.1,
|
| 134 |
+
"weight_decay_encoder": 0.01,
|
| 135 |
+
"weight_decay_other": 0.01,
|
| 136 |
+
"words_splitter_type": "whitespace"
|
| 137 |
+
}
|
checkpoint-50000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a08508dc627ac96d10b890f291dda9b30f5cfccca0d67252e992a338caf7489f
|
| 3 |
+
size 3522103130
|
checkpoint-50000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca93e63f3dacfc37f396280777a672fd5761a0850e12071b4a894f77d95a0623
|
| 3 |
+
size 1761030542
|
checkpoint-50000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9256c7b6f93ee9100b4c452f445bd8772c2feac8e2abcbb69e96e6279096bc52
|
| 3 |
+
size 14244
|
checkpoint-50000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a2a02082cf5795dc302564c2c80d261353663a200b6a06d5959ba07ef981c9f
|
| 3 |
+
size 1064
|
checkpoint-50000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"eos_token": "[SEP]",
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"pad_token": "[PAD]",
|
| 7 |
+
"sep_token": "[SEP]",
|
| 8 |
+
"unk_token": {
|
| 9 |
+
"content": "[UNK]",
|
| 10 |
+
"lstrip": false,
|
| 11 |
+
"normalized": true,
|
| 12 |
+
"rstrip": false,
|
| 13 |
+
"single_word": false
|
| 14 |
+
}
|
| 15 |
+
}
|
checkpoint-50000/spm.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
| 3 |
+
size 2464616
|
checkpoint-50000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-50000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "[CLS]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[SEP]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[UNK]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"128000": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"128001": {
|
| 44 |
+
"content": "[FLERT]",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
"128002": {
|
| 52 |
+
"content": "<<ENT>>",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
+
},
|
| 59 |
+
"128003": {
|
| 60 |
+
"content": "<<SEP>>",
|
| 61 |
+
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
+
}
|
| 67 |
+
},
|
| 68 |
+
"bos_token": "[CLS]",
|
| 69 |
+
"clean_up_tokenization_spaces": true,
|
| 70 |
+
"cls_token": "[CLS]",
|
| 71 |
+
"do_lower_case": false,
|
| 72 |
+
"eos_token": "[SEP]",
|
| 73 |
+
"mask_token": "[MASK]",
|
| 74 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 75 |
+
"pad_token": "[PAD]",
|
| 76 |
+
"sep_token": "[SEP]",
|
| 77 |
+
"sp_model_kwargs": {},
|
| 78 |
+
"split_by_punct": false,
|
| 79 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 80 |
+
"unk_token": "[UNK]",
|
| 81 |
+
"vocab_type": "spm"
|
| 82 |
+
}
|
checkpoint-50000/trainer_state.json
ADDED
|
@@ -0,0 +1,378 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3657457189463597,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 50000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 7.314914378927195e-06,
|
| 13 |
+
"grad_norm": 2702.876220703125,
|
| 14 |
+
"learning_rate": 3.6573769292663306e-10,
|
| 15 |
+
"loss": 185.8854,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.007314914378927195,
|
| 20 |
+
"grad_norm": 25910.705078125,
|
| 21 |
+
"learning_rate": 3.6573769292663306e-07,
|
| 22 |
+
"loss": 48.434,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01462982875785439,
|
| 27 |
+
"grad_norm": 3.9086620807647705,
|
| 28 |
+
"learning_rate": 7.314753858532661e-07,
|
| 29 |
+
"loss": 17.9804,
|
| 30 |
+
"step": 2000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.021944743136781585,
|
| 34 |
+
"grad_norm": 17.556304931640625,
|
| 35 |
+
"learning_rate": 1.097213078779899e-06,
|
| 36 |
+
"loss": 10.5262,
|
| 37 |
+
"step": 3000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02925965751570878,
|
| 41 |
+
"grad_norm": 4689.71240234375,
|
| 42 |
+
"learning_rate": 1.4629507717065323e-06,
|
| 43 |
+
"loss": 11.6651,
|
| 44 |
+
"step": 4000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03657457189463598,
|
| 48 |
+
"grad_norm": 0.0065016308799386024,
|
| 49 |
+
"learning_rate": 1.8286884646331652e-06,
|
| 50 |
+
"loss": 9.4746,
|
| 51 |
+
"step": 5000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04388948627356317,
|
| 55 |
+
"grad_norm": 0.060599055141210556,
|
| 56 |
+
"learning_rate": 2.194426157559798e-06,
|
| 57 |
+
"loss": 7.1303,
|
| 58 |
+
"step": 6000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.051204400652490364,
|
| 62 |
+
"grad_norm": 1028.9453125,
|
| 63 |
+
"learning_rate": 2.560163850486431e-06,
|
| 64 |
+
"loss": 8.917,
|
| 65 |
+
"step": 7000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05851931503141756,
|
| 69 |
+
"grad_norm": 5.463726043701172,
|
| 70 |
+
"learning_rate": 2.9259015434130645e-06,
|
| 71 |
+
"loss": 8.9339,
|
| 72 |
+
"step": 8000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06583422941034475,
|
| 76 |
+
"grad_norm": 0.3307730257511139,
|
| 77 |
+
"learning_rate": 3.2916392363396975e-06,
|
| 78 |
+
"loss": 8.746,
|
| 79 |
+
"step": 9000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07314914378927195,
|
| 83 |
+
"grad_norm": 0.003347629914060235,
|
| 84 |
+
"learning_rate": 3.6573769292663304e-06,
|
| 85 |
+
"loss": 9.4711,
|
| 86 |
+
"step": 10000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08046405816819914,
|
| 90 |
+
"grad_norm": 0.006155295763164759,
|
| 91 |
+
"learning_rate": 4.023114622192964e-06,
|
| 92 |
+
"loss": 7.647,
|
| 93 |
+
"step": 11000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08777897254712634,
|
| 97 |
+
"grad_norm": 0.0019895241130143404,
|
| 98 |
+
"learning_rate": 4.388852315119596e-06,
|
| 99 |
+
"loss": 5.7327,
|
| 100 |
+
"step": 12000
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.09509388692605353,
|
| 104 |
+
"grad_norm": 0.004527593031525612,
|
| 105 |
+
"learning_rate": 4.75459000804623e-06,
|
| 106 |
+
"loss": 5.5751,
|
| 107 |
+
"step": 13000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.10240880130498073,
|
| 111 |
+
"grad_norm": 0.009128883481025696,
|
| 112 |
+
"learning_rate": 4.986629929451543e-06,
|
| 113 |
+
"loss": 8.6148,
|
| 114 |
+
"step": 14000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.10972371568390792,
|
| 118 |
+
"grad_norm": 0.012683026492595673,
|
| 119 |
+
"learning_rate": 4.9459914171462015e-06,
|
| 120 |
+
"loss": 8.5558,
|
| 121 |
+
"step": 15000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11703863006283512,
|
| 125 |
+
"grad_norm": 0.0028167981654405594,
|
| 126 |
+
"learning_rate": 4.90535290484086e-06,
|
| 127 |
+
"loss": 6.0433,
|
| 128 |
+
"step": 16000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1243535444417623,
|
| 132 |
+
"grad_norm": 0.0011900264071300626,
|
| 133 |
+
"learning_rate": 4.864714392535519e-06,
|
| 134 |
+
"loss": 6.9084,
|
| 135 |
+
"step": 17000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1316684588206895,
|
| 139 |
+
"grad_norm": 0.025524910539388657,
|
| 140 |
+
"learning_rate": 4.824075880230177e-06,
|
| 141 |
+
"loss": 6.7333,
|
| 142 |
+
"step": 18000
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1389833731996167,
|
| 146 |
+
"grad_norm": 0.027619725093245506,
|
| 147 |
+
"learning_rate": 4.783437367924835e-06,
|
| 148 |
+
"loss": 4.1436,
|
| 149 |
+
"step": 19000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1462982875785439,
|
| 153 |
+
"grad_norm": 0.012698939070105553,
|
| 154 |
+
"learning_rate": 4.742798855619494e-06,
|
| 155 |
+
"loss": 5.397,
|
| 156 |
+
"step": 20000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.15361320195747108,
|
| 160 |
+
"grad_norm": 0.031142177060246468,
|
| 161 |
+
"learning_rate": 4.702160343314152e-06,
|
| 162 |
+
"loss": 5.5156,
|
| 163 |
+
"step": 21000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.16092811633639828,
|
| 167 |
+
"grad_norm": 0.0004189134924672544,
|
| 168 |
+
"learning_rate": 4.66152183100881e-06,
|
| 169 |
+
"loss": 4.8633,
|
| 170 |
+
"step": 22000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.16824303071532548,
|
| 174 |
+
"grad_norm": 0.0103899035602808,
|
| 175 |
+
"learning_rate": 4.620883318703469e-06,
|
| 176 |
+
"loss": 7.2146,
|
| 177 |
+
"step": 23000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17555794509425268,
|
| 181 |
+
"grad_norm": 1158.8017578125,
|
| 182 |
+
"learning_rate": 4.580244806398127e-06,
|
| 183 |
+
"loss": 5.7667,
|
| 184 |
+
"step": 24000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18287285947317986,
|
| 188 |
+
"grad_norm": 0.03399639576673508,
|
| 189 |
+
"learning_rate": 4.5396062940927856e-06,
|
| 190 |
+
"loss": 5.0472,
|
| 191 |
+
"step": 25000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.19018777385210706,
|
| 195 |
+
"grad_norm": 0.017644532024860382,
|
| 196 |
+
"learning_rate": 4.4989677817874446e-06,
|
| 197 |
+
"loss": 4.8188,
|
| 198 |
+
"step": 26000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.19750268823103426,
|
| 202 |
+
"grad_norm": 0.00010079160711029544,
|
| 203 |
+
"learning_rate": 4.4583292694821035e-06,
|
| 204 |
+
"loss": 5.7598,
|
| 205 |
+
"step": 27000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.20481760260996146,
|
| 209 |
+
"grad_norm": 0.009618501178920269,
|
| 210 |
+
"learning_rate": 4.417690757176762e-06,
|
| 211 |
+
"loss": 4.683,
|
| 212 |
+
"step": 28000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.21213251698888866,
|
| 216 |
+
"grad_norm": 0.018396975472569466,
|
| 217 |
+
"learning_rate": 4.377052244871421e-06,
|
| 218 |
+
"loss": 5.7816,
|
| 219 |
+
"step": 29000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.21944743136781583,
|
| 223 |
+
"grad_norm": 0.026549218222498894,
|
| 224 |
+
"learning_rate": 4.336413732566079e-06,
|
| 225 |
+
"loss": 5.5149,
|
| 226 |
+
"step": 30000
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.22676234574674303,
|
| 230 |
+
"grad_norm": 0.01402178592979908,
|
| 231 |
+
"learning_rate": 4.295775220260737e-06,
|
| 232 |
+
"loss": 6.1021,
|
| 233 |
+
"step": 31000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.23407726012567023,
|
| 237 |
+
"grad_norm": 2950.190185546875,
|
| 238 |
+
"learning_rate": 4.255136707955396e-06,
|
| 239 |
+
"loss": 5.1742,
|
| 240 |
+
"step": 32000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.24139217450459743,
|
| 244 |
+
"grad_norm": 0.01243713591247797,
|
| 245 |
+
"learning_rate": 4.214498195650054e-06,
|
| 246 |
+
"loss": 4.8856,
|
| 247 |
+
"step": 33000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2487070888835246,
|
| 251 |
+
"grad_norm": 0.00121857482008636,
|
| 252 |
+
"learning_rate": 4.173859683344712e-06,
|
| 253 |
+
"loss": 4.0296,
|
| 254 |
+
"step": 34000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.25602200326245184,
|
| 258 |
+
"grad_norm": 0.021528728306293488,
|
| 259 |
+
"learning_rate": 4.133221171039371e-06,
|
| 260 |
+
"loss": 3.7989,
|
| 261 |
+
"step": 35000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.263336917641379,
|
| 265 |
+
"grad_norm": 0.067794568836689,
|
| 266 |
+
"learning_rate": 4.092582658734029e-06,
|
| 267 |
+
"loss": 4.8373,
|
| 268 |
+
"step": 36000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2706518320203062,
|
| 272 |
+
"grad_norm": 83.66200256347656,
|
| 273 |
+
"learning_rate": 4.0519441464286876e-06,
|
| 274 |
+
"loss": 3.2441,
|
| 275 |
+
"step": 37000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.2779667463992334,
|
| 279 |
+
"grad_norm": 0.00217541866004467,
|
| 280 |
+
"learning_rate": 4.0113056341233466e-06,
|
| 281 |
+
"loss": 3.5578,
|
| 282 |
+
"step": 38000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2852816607781606,
|
| 286 |
+
"grad_norm": 0.0008728219545446336,
|
| 287 |
+
"learning_rate": 3.970667121818005e-06,
|
| 288 |
+
"loss": 2.6644,
|
| 289 |
+
"step": 39000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2925965751570878,
|
| 293 |
+
"grad_norm": 0.027021408081054688,
|
| 294 |
+
"learning_rate": 3.930028609512664e-06,
|
| 295 |
+
"loss": 3.7778,
|
| 296 |
+
"step": 40000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.299911489536015,
|
| 300 |
+
"grad_norm": 0.001312136766500771,
|
| 301 |
+
"learning_rate": 3.889390097207322e-06,
|
| 302 |
+
"loss": 4.2509,
|
| 303 |
+
"step": 41000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.30722640391494216,
|
| 307 |
+
"grad_norm": 0.011145360767841339,
|
| 308 |
+
"learning_rate": 3.84875158490198e-06,
|
| 309 |
+
"loss": 4.6084,
|
| 310 |
+
"step": 42000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3145413182938694,
|
| 314 |
+
"grad_norm": 289.0566711425781,
|
| 315 |
+
"learning_rate": 3.8081130725966386e-06,
|
| 316 |
+
"loss": 3.3176,
|
| 317 |
+
"step": 43000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.32185623267279656,
|
| 321 |
+
"grad_norm": 0.006871068850159645,
|
| 322 |
+
"learning_rate": 3.767474560291297e-06,
|
| 323 |
+
"loss": 3.0108,
|
| 324 |
+
"step": 44000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.32917114705172373,
|
| 328 |
+
"grad_norm": 0.002771923318505287,
|
| 329 |
+
"learning_rate": 3.7268360479859557e-06,
|
| 330 |
+
"loss": 3.5366,
|
| 331 |
+
"step": 45000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.33648606143065096,
|
| 335 |
+
"grad_norm": 0.0030192858539521694,
|
| 336 |
+
"learning_rate": 3.686197535680614e-06,
|
| 337 |
+
"loss": 3.3997,
|
| 338 |
+
"step": 46000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.34380097580957814,
|
| 342 |
+
"grad_norm": 3926.616943359375,
|
| 343 |
+
"learning_rate": 3.6455590233752724e-06,
|
| 344 |
+
"loss": 3.5042,
|
| 345 |
+
"step": 47000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.35111589018850536,
|
| 349 |
+
"grad_norm": 0.000544128124602139,
|
| 350 |
+
"learning_rate": 3.604920511069931e-06,
|
| 351 |
+
"loss": 2.9953,
|
| 352 |
+
"step": 48000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.35843080456743254,
|
| 356 |
+
"grad_norm": 14144.9677734375,
|
| 357 |
+
"learning_rate": 3.564281998764589e-06,
|
| 358 |
+
"loss": 2.9742,
|
| 359 |
+
"step": 49000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.3657457189463597,
|
| 363 |
+
"grad_norm": 0.0021343908738344908,
|
| 364 |
+
"learning_rate": 3.5236434864592477e-06,
|
| 365 |
+
"loss": 3.4145,
|
| 366 |
+
"step": 50000
|
| 367 |
+
}
|
| 368 |
+
],
|
| 369 |
+
"logging_steps": 1000,
|
| 370 |
+
"max_steps": 136707,
|
| 371 |
+
"num_input_tokens_seen": 0,
|
| 372 |
+
"num_train_epochs": 1,
|
| 373 |
+
"save_steps": 50000,
|
| 374 |
+
"total_flos": 0.0,
|
| 375 |
+
"train_batch_size": 2,
|
| 376 |
+
"trial_name": null,
|
| 377 |
+
"trial_params": null
|
| 378 |
+
}
|
gliner_config.json
ADDED
|
@@ -0,0 +1,137 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"class_token_index": 128002,
|
| 3 |
+
"dropout": 0.1,
|
| 4 |
+
"embed_ent_token": true,
|
| 5 |
+
"encoder_config": {
|
| 6 |
+
"_name_or_path": "microsoft/deberta-v3-large",
|
| 7 |
+
"add_cross_attention": false,
|
| 8 |
+
"architectures": null,
|
| 9 |
+
"attention_probs_dropout_prob": 0.1,
|
| 10 |
+
"bad_words_ids": null,
|
| 11 |
+
"begin_suppress_tokens": null,
|
| 12 |
+
"bos_token_id": null,
|
| 13 |
+
"chunk_size_feed_forward": 0,
|
| 14 |
+
"cross_attention_hidden_size": null,
|
| 15 |
+
"decoder_start_token_id": null,
|
| 16 |
+
"diversity_penalty": 0.0,
|
| 17 |
+
"do_sample": false,
|
| 18 |
+
"early_stopping": false,
|
| 19 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 20 |
+
"eos_token_id": null,
|
| 21 |
+
"exponential_decay_length_penalty": null,
|
| 22 |
+
"finetuning_task": null,
|
| 23 |
+
"forced_bos_token_id": null,
|
| 24 |
+
"forced_eos_token_id": null,
|
| 25 |
+
"hidden_act": "gelu",
|
| 26 |
+
"hidden_dropout_prob": 0.1,
|
| 27 |
+
"hidden_size": 1024,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "LABEL_0",
|
| 30 |
+
"1": "LABEL_1"
|
| 31 |
+
},
|
| 32 |
+
"initializer_range": 0.02,
|
| 33 |
+
"intermediate_size": 4096,
|
| 34 |
+
"is_decoder": false,
|
| 35 |
+
"is_encoder_decoder": false,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1
|
| 39 |
+
},
|
| 40 |
+
"layer_norm_eps": 1e-07,
|
| 41 |
+
"length_penalty": 1.0,
|
| 42 |
+
"max_length": 20,
|
| 43 |
+
"max_position_embeddings": 512,
|
| 44 |
+
"max_relative_positions": -1,
|
| 45 |
+
"min_length": 0,
|
| 46 |
+
"model_type": "deberta-v2",
|
| 47 |
+
"no_repeat_ngram_size": 0,
|
| 48 |
+
"norm_rel_ebd": "layer_norm",
|
| 49 |
+
"num_attention_heads": 16,
|
| 50 |
+
"num_beam_groups": 1,
|
| 51 |
+
"num_beams": 1,
|
| 52 |
+
"num_hidden_layers": 24,
|
| 53 |
+
"num_return_sequences": 1,
|
| 54 |
+
"output_attentions": false,
|
| 55 |
+
"output_hidden_states": false,
|
| 56 |
+
"output_scores": false,
|
| 57 |
+
"pad_token_id": 0,
|
| 58 |
+
"pooler_dropout": 0,
|
| 59 |
+
"pooler_hidden_act": "gelu",
|
| 60 |
+
"pooler_hidden_size": 1024,
|
| 61 |
+
"pos_att_type": [
|
| 62 |
+
"p2c",
|
| 63 |
+
"c2p"
|
| 64 |
+
],
|
| 65 |
+
"position_biased_input": false,
|
| 66 |
+
"position_buckets": 256,
|
| 67 |
+
"prefix": null,
|
| 68 |
+
"problem_type": null,
|
| 69 |
+
"pruned_heads": {},
|
| 70 |
+
"relative_attention": true,
|
| 71 |
+
"remove_invalid_values": false,
|
| 72 |
+
"repetition_penalty": 1.0,
|
| 73 |
+
"return_dict": true,
|
| 74 |
+
"return_dict_in_generate": false,
|
| 75 |
+
"sep_token_id": null,
|
| 76 |
+
"share_att_key": true,
|
| 77 |
+
"suppress_tokens": null,
|
| 78 |
+
"task_specific_params": null,
|
| 79 |
+
"temperature": 1.0,
|
| 80 |
+
"tf_legacy_loss": false,
|
| 81 |
+
"tie_encoder_decoder": false,
|
| 82 |
+
"tie_word_embeddings": true,
|
| 83 |
+
"tokenizer_class": null,
|
| 84 |
+
"top_k": 50,
|
| 85 |
+
"top_p": 1.0,
|
| 86 |
+
"torch_dtype": null,
|
| 87 |
+
"torchscript": false,
|
| 88 |
+
"type_vocab_size": 0,
|
| 89 |
+
"typical_p": 1.0,
|
| 90 |
+
"use_bfloat16": false,
|
| 91 |
+
"vocab_size": 128004
|
| 92 |
+
},
|
| 93 |
+
"ent_token": "<<ENT>>",
|
| 94 |
+
"eval_every": 1000,
|
| 95 |
+
"fine_tune": true,
|
| 96 |
+
"freeze_token_rep": false,
|
| 97 |
+
"fuse_layers": false,
|
| 98 |
+
"has_rnn": true,
|
| 99 |
+
"hidden_size": 512,
|
| 100 |
+
"labels_encoder": null,
|
| 101 |
+
"labels_encoder_config": null,
|
| 102 |
+
"log_dir": "logs_final",
|
| 103 |
+
"loss_alpha": 0.75,
|
| 104 |
+
"loss_gamma": 0,
|
| 105 |
+
"loss_reduction": "sum",
|
| 106 |
+
"lr_encoder": "5e-6",
|
| 107 |
+
"lr_others": "7e-6",
|
| 108 |
+
"max_len": 768,
|
| 109 |
+
"max_neg_type_ratio": 1,
|
| 110 |
+
"max_types": 30,
|
| 111 |
+
"max_width": 100,
|
| 112 |
+
"model_name": "microsoft/deberta-v3-large",
|
| 113 |
+
"model_type": "gliner",
|
| 114 |
+
"name": "token level gliner large",
|
| 115 |
+
"num_post_fusion_layers": 1,
|
| 116 |
+
"num_steps": 6000,
|
| 117 |
+
"post_fusion_schema": "",
|
| 118 |
+
"prev_path": "logs_large/model_120000",
|
| 119 |
+
"random_drop": true,
|
| 120 |
+
"root_dir": "gliner_logs",
|
| 121 |
+
"save_total_limit": 10,
|
| 122 |
+
"scheduler_type": "linear",
|
| 123 |
+
"sep_token": "<<SEP>>",
|
| 124 |
+
"shuffle_types": true,
|
| 125 |
+
"size_sup": -1,
|
| 126 |
+
"span_mode": "token_level",
|
| 127 |
+
"subtoken_pooling": "first",
|
| 128 |
+
"train_batch_size": 8,
|
| 129 |
+
"train_data": "../../data/unie_ner1.json",
|
| 130 |
+
"transformers_version": "4.40.1",
|
| 131 |
+
"val_data_dir": "none",
|
| 132 |
+
"vocab_size": 128004,
|
| 133 |
+
"warmup_ratio": 0.1,
|
| 134 |
+
"weight_decay_encoder": 0.01,
|
| 135 |
+
"weight_decay_other": 0.01,
|
| 136 |
+
"words_splitter_type": "whitespace"
|
| 137 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4afe5473217129bf534b8e7e72b0fea59e7d58a291541347898b438c5bdb5838
|
| 3 |
+
size 1761030542
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"eos_token": "[SEP]",
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"pad_token": "[PAD]",
|
| 7 |
+
"sep_token": "[SEP]",
|
| 8 |
+
"unk_token": {
|
| 9 |
+
"content": "[UNK]",
|
| 10 |
+
"lstrip": false,
|
| 11 |
+
"normalized": true,
|
| 12 |
+
"rstrip": false,
|
| 13 |
+
"single_word": false
|
| 14 |
+
}
|
| 15 |
+
}
|
spm.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
| 3 |
+
size 2464616
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "[CLS]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[SEP]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[UNK]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"128000": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"128001": {
|
| 44 |
+
"content": "[FLERT]",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
"128002": {
|
| 52 |
+
"content": "<<ENT>>",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
+
},
|
| 59 |
+
"128003": {
|
| 60 |
+
"content": "<<SEP>>",
|
| 61 |
+
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
+
"rstrip": false,
|
| 64 |
+
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
+
}
|
| 67 |
+
},
|
| 68 |
+
"bos_token": "[CLS]",
|
| 69 |
+
"clean_up_tokenization_spaces": true,
|
| 70 |
+
"cls_token": "[CLS]",
|
| 71 |
+
"do_lower_case": false,
|
| 72 |
+
"eos_token": "[SEP]",
|
| 73 |
+
"mask_token": "[MASK]",
|
| 74 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 75 |
+
"pad_token": "[PAD]",
|
| 76 |
+
"sep_token": "[SEP]",
|
| 77 |
+
"sp_model_kwargs": {},
|
| 78 |
+
"split_by_punct": false,
|
| 79 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 80 |
+
"unk_token": "[UNK]",
|
| 81 |
+
"vocab_type": "spm"
|
| 82 |
+
}
|