Training in progress, step 5500, checkpoint
Browse files- checkpoint-5500/config.json +30 -0
- checkpoint-5500/generation_config.json +6 -0
- checkpoint-5500/model.safetensors +3 -0
- checkpoint-5500/optimizer.pt +3 -0
- checkpoint-5500/rng_state.pth +3 -0
- checkpoint-5500/scheduler.pt +3 -0
- checkpoint-5500/special_tokens_map.json +24 -0
- checkpoint-5500/tokenizer.json +96 -0
- checkpoint-5500/tokenizer_config.json +38 -0
- checkpoint-5500/trainer_state.json +418 -0
- checkpoint-5500/training_args.bin +3 -0
checkpoint-5500/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/pythia-160m",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTNeoXForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": true,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"classifier_dropout": 0.1,
|
| 10 |
+
"eos_token_id": 0,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout": 0.0,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"layer_norm_eps": 1e-05,
|
| 17 |
+
"max_position_embeddings": 2048,
|
| 18 |
+
"model_type": "gpt_neox",
|
| 19 |
+
"num_attention_heads": 12,
|
| 20 |
+
"num_hidden_layers": 12,
|
| 21 |
+
"rope_scaling": null,
|
| 22 |
+
"rotary_emb_base": 10000,
|
| 23 |
+
"rotary_pct": 0.25,
|
| 24 |
+
"tie_word_embeddings": false,
|
| 25 |
+
"torch_dtype": "float32",
|
| 26 |
+
"transformers_version": "4.42.3",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"use_parallel_residual": true,
|
| 29 |
+
"vocab_size": 8
|
| 30 |
+
}
|
checkpoint-5500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.42.3"
|
| 6 |
+
}
|
checkpoint-5500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86646750ed9f5662722a1f1e68c12bf006061420b80114ed6f5a7ed29d9f0302
|
| 3 |
+
size 340289960
|
checkpoint-5500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db4a28de97ffb97fe2c3265042c1bfb101ff28e45ed765f99b972da91616fbc6
|
| 3 |
+
size 680670138
|
checkpoint-5500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:835f869ea325fd6edf27b48b589309fb66641cb92b45f2fc13d1bb6e8814106c
|
| 3 |
+
size 14244
|
checkpoint-5500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:791717f94f4cfe85f1c44386b0235a5d9d1d9f583641bfaaeabf6a2238ba9814
|
| 3 |
+
size 1064
|
checkpoint-5500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|beginoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|endoftext|>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<|unknown|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-5500/tokenizer.json
ADDED
|
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
| 5 |
+
"added_tokens": [
|
| 6 |
+
{
|
| 7 |
+
"id": 0,
|
| 8 |
+
"content": "<|beginoftext|>",
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"lstrip": false,
|
| 11 |
+
"rstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"special": true
|
| 14 |
+
},
|
| 15 |
+
{
|
| 16 |
+
"id": 1,
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"normalized": false,
|
| 22 |
+
"special": true
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"id": 2,
|
| 26 |
+
"content": "<|unknown|>",
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"lstrip": false,
|
| 29 |
+
"rstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"special": true
|
| 32 |
+
}
|
| 33 |
+
],
|
| 34 |
+
"normalizer": {
|
| 35 |
+
"type": "NFC"
|
| 36 |
+
},
|
| 37 |
+
"pre_tokenizer": {
|
| 38 |
+
"type": "ByteLevel",
|
| 39 |
+
"add_prefix_space": false,
|
| 40 |
+
"trim_offsets": true,
|
| 41 |
+
"use_regex": true
|
| 42 |
+
},
|
| 43 |
+
"post_processor": {
|
| 44 |
+
"type": "TemplateProcessing",
|
| 45 |
+
"single": [
|
| 46 |
+
{
|
| 47 |
+
"Sequence": {
|
| 48 |
+
"id": "A",
|
| 49 |
+
"type_id": 0
|
| 50 |
+
}
|
| 51 |
+
}
|
| 52 |
+
],
|
| 53 |
+
"pair": [
|
| 54 |
+
{
|
| 55 |
+
"Sequence": {
|
| 56 |
+
"id": "A",
|
| 57 |
+
"type_id": 0
|
| 58 |
+
}
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"Sequence": {
|
| 62 |
+
"id": "B",
|
| 63 |
+
"type_id": 1
|
| 64 |
+
}
|
| 65 |
+
}
|
| 66 |
+
],
|
| 67 |
+
"special_tokens": {}
|
| 68 |
+
},
|
| 69 |
+
"decoder": {
|
| 70 |
+
"type": "ByteLevel",
|
| 71 |
+
"add_prefix_space": false,
|
| 72 |
+
"trim_offsets": true,
|
| 73 |
+
"use_regex": true
|
| 74 |
+
},
|
| 75 |
+
"model": {
|
| 76 |
+
"type": "BPE",
|
| 77 |
+
"dropout": null,
|
| 78 |
+
"unk_token": "<|unknown|>",
|
| 79 |
+
"continuing_subword_prefix": null,
|
| 80 |
+
"end_of_word_suffix": null,
|
| 81 |
+
"fuse_unk": false,
|
| 82 |
+
"byte_fallback": false,
|
| 83 |
+
"ignore_merges": false,
|
| 84 |
+
"vocab": {
|
| 85 |
+
"<|beginoftext|>": 0,
|
| 86 |
+
"<|padding|>": 1,
|
| 87 |
+
"<|unknown|>": 2,
|
| 88 |
+
"a": 3,
|
| 89 |
+
"c": 4,
|
| 90 |
+
"g": 5,
|
| 91 |
+
"t": 6,
|
| 92 |
+
"n": 7
|
| 93 |
+
},
|
| 94 |
+
"merges": []
|
| 95 |
+
}
|
| 96 |
+
}
|
checkpoint-5500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": false,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<|beginoftext|>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": true,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<|endoftext|>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "<|unknown|>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"bos_token": "<|beginoftext|>",
|
| 32 |
+
"clean_up_tokenization_spaces": true,
|
| 33 |
+
"eos_token": "<|endoftext|>",
|
| 34 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 35 |
+
"pad_token": "<|endoftext|>",
|
| 36 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 37 |
+
"unk_token": "<|unknown|>"
|
| 38 |
+
}
|
checkpoint-5500/trainer_state.json
ADDED
|
@@ -0,0 +1,418 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.05374505301216593,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 5500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0009771827820393806,
|
| 13 |
+
"grad_norm": 0.5417118072509766,
|
| 14 |
+
"learning_rate": 4.995602247740044e-05,
|
| 15 |
+
"loss": 1.378,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.001954365564078761,
|
| 20 |
+
"grad_norm": 0.6493918895721436,
|
| 21 |
+
"learning_rate": 4.990715856340093e-05,
|
| 22 |
+
"loss": 1.3304,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0029315483461181415,
|
| 27 |
+
"grad_norm": 0.9062462449073792,
|
| 28 |
+
"learning_rate": 4.9858294649401425e-05,
|
| 29 |
+
"loss": 1.3284,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.003908731128157522,
|
| 34 |
+
"grad_norm": 0.750052273273468,
|
| 35 |
+
"learning_rate": 4.9809430735401906e-05,
|
| 36 |
+
"loss": 1.3166,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.004885913910196903,
|
| 41 |
+
"grad_norm": 0.6602022051811218,
|
| 42 |
+
"learning_rate": 4.97605668214024e-05,
|
| 43 |
+
"loss": 1.3166,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.005863096692236283,
|
| 48 |
+
"grad_norm": 0.4193927049636841,
|
| 49 |
+
"learning_rate": 4.971170290740288e-05,
|
| 50 |
+
"loss": 1.3098,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.006840279474275663,
|
| 55 |
+
"grad_norm": 0.6095415949821472,
|
| 56 |
+
"learning_rate": 4.966283899340338e-05,
|
| 57 |
+
"loss": 1.3103,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.007817462256315045,
|
| 62 |
+
"grad_norm": 0.9943467378616333,
|
| 63 |
+
"learning_rate": 4.9613975079403865e-05,
|
| 64 |
+
"loss": 1.3096,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.008794645038354424,
|
| 69 |
+
"grad_norm": 1.2263585329055786,
|
| 70 |
+
"learning_rate": 4.9565111165404346e-05,
|
| 71 |
+
"loss": 1.3067,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.009771827820393805,
|
| 76 |
+
"grad_norm": 0.7198677659034729,
|
| 77 |
+
"learning_rate": 4.951624725140484e-05,
|
| 78 |
+
"loss": 1.3041,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.010749010602433185,
|
| 83 |
+
"grad_norm": 0.7370775938034058,
|
| 84 |
+
"learning_rate": 4.946738333740533e-05,
|
| 85 |
+
"loss": 1.302,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.011726193384472566,
|
| 90 |
+
"grad_norm": 0.5109437704086304,
|
| 91 |
+
"learning_rate": 4.941851942340582e-05,
|
| 92 |
+
"loss": 1.3089,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.012703376166511945,
|
| 97 |
+
"grad_norm": 0.1879555583000183,
|
| 98 |
+
"learning_rate": 4.9369655509406305e-05,
|
| 99 |
+
"loss": 1.3043,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.013680558948551327,
|
| 104 |
+
"grad_norm": 0.951046884059906,
|
| 105 |
+
"learning_rate": 4.932079159540679e-05,
|
| 106 |
+
"loss": 1.3098,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.014657741730590706,
|
| 111 |
+
"grad_norm": 0.2478829026222229,
|
| 112 |
+
"learning_rate": 4.927192768140728e-05,
|
| 113 |
+
"loss": 1.3026,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.01563492451263009,
|
| 118 |
+
"grad_norm": 0.5585843324661255,
|
| 119 |
+
"learning_rate": 4.9223063767407776e-05,
|
| 120 |
+
"loss": 1.3014,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.016612107294669467,
|
| 125 |
+
"grad_norm": 0.48532453179359436,
|
| 126 |
+
"learning_rate": 4.917419985340826e-05,
|
| 127 |
+
"loss": 1.2981,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.017589290076708848,
|
| 132 |
+
"grad_norm": 0.4233573079109192,
|
| 133 |
+
"learning_rate": 4.912533593940875e-05,
|
| 134 |
+
"loss": 1.2992,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.01856647285874823,
|
| 139 |
+
"grad_norm": 0.3272475600242615,
|
| 140 |
+
"learning_rate": 4.9076472025409234e-05,
|
| 141 |
+
"loss": 1.292,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.01954365564078761,
|
| 146 |
+
"grad_norm": 0.5299385786056519,
|
| 147 |
+
"learning_rate": 4.902760811140973e-05,
|
| 148 |
+
"loss": 1.2963,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.02052083842282699,
|
| 153 |
+
"grad_norm": 0.1614024043083191,
|
| 154 |
+
"learning_rate": 4.8978744197410216e-05,
|
| 155 |
+
"loss": 1.2945,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.02149802120486637,
|
| 160 |
+
"grad_norm": 0.6039963960647583,
|
| 161 |
+
"learning_rate": 4.8929880283410705e-05,
|
| 162 |
+
"loss": 1.2913,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.02247520398690575,
|
| 167 |
+
"grad_norm": 0.5772804021835327,
|
| 168 |
+
"learning_rate": 4.888101636941119e-05,
|
| 169 |
+
"loss": 1.2895,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.023452386768945132,
|
| 174 |
+
"grad_norm": 0.7489622235298157,
|
| 175 |
+
"learning_rate": 4.883215245541168e-05,
|
| 176 |
+
"loss": 1.2847,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.024429569550984513,
|
| 181 |
+
"grad_norm": 0.30208253860473633,
|
| 182 |
+
"learning_rate": 4.878328854141217e-05,
|
| 183 |
+
"loss": 1.2924,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.02540675233302389,
|
| 188 |
+
"grad_norm": 0.36944472789764404,
|
| 189 |
+
"learning_rate": 4.873442462741266e-05,
|
| 190 |
+
"loss": 1.2916,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.026383935115063272,
|
| 195 |
+
"grad_norm": 0.3268676698207855,
|
| 196 |
+
"learning_rate": 4.8685560713413145e-05,
|
| 197 |
+
"loss": 1.2893,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.027361117897102653,
|
| 202 |
+
"grad_norm": 0.2795974910259247,
|
| 203 |
+
"learning_rate": 4.863669679941363e-05,
|
| 204 |
+
"loss": 1.282,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.028338300679142035,
|
| 209 |
+
"grad_norm": 0.36298853158950806,
|
| 210 |
+
"learning_rate": 4.858783288541413e-05,
|
| 211 |
+
"loss": 1.2832,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.029315483461181412,
|
| 216 |
+
"grad_norm": 0.5242423415184021,
|
| 217 |
+
"learning_rate": 4.853896897141461e-05,
|
| 218 |
+
"loss": 1.2819,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.030292666243220794,
|
| 223 |
+
"grad_norm": 0.25340864062309265,
|
| 224 |
+
"learning_rate": 4.8490105057415104e-05,
|
| 225 |
+
"loss": 1.2809,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.03126984902526018,
|
| 230 |
+
"grad_norm": 0.7241976261138916,
|
| 231 |
+
"learning_rate": 4.844124114341559e-05,
|
| 232 |
+
"loss": 1.2802,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.032247031807299556,
|
| 237 |
+
"grad_norm": 0.5154001712799072,
|
| 238 |
+
"learning_rate": 4.839237722941608e-05,
|
| 239 |
+
"loss": 1.2748,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.033224214589338934,
|
| 244 |
+
"grad_norm": 0.5323473811149597,
|
| 245 |
+
"learning_rate": 4.834351331541657e-05,
|
| 246 |
+
"loss": 1.284,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.03420139737137832,
|
| 251 |
+
"grad_norm": 0.3947168290615082,
|
| 252 |
+
"learning_rate": 4.8294649401417056e-05,
|
| 253 |
+
"loss": 1.276,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.035178580153417696,
|
| 258 |
+
"grad_norm": 0.4776057302951813,
|
| 259 |
+
"learning_rate": 4.8245785487417544e-05,
|
| 260 |
+
"loss": 1.2783,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.036155762935457074,
|
| 265 |
+
"grad_norm": 0.4884164035320282,
|
| 266 |
+
"learning_rate": 4.819692157341804e-05,
|
| 267 |
+
"loss": 1.2745,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.03713294571749646,
|
| 272 |
+
"grad_norm": 0.5210428833961487,
|
| 273 |
+
"learning_rate": 4.814805765941852e-05,
|
| 274 |
+
"loss": 1.2707,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.038110128499535836,
|
| 279 |
+
"grad_norm": 0.46214359998703003,
|
| 280 |
+
"learning_rate": 4.809919374541901e-05,
|
| 281 |
+
"loss": 1.2727,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.03908731128157522,
|
| 286 |
+
"grad_norm": 0.2656782865524292,
|
| 287 |
+
"learning_rate": 4.8050329831419496e-05,
|
| 288 |
+
"loss": 1.2694,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.0400644940636146,
|
| 293 |
+
"grad_norm": 0.4923059940338135,
|
| 294 |
+
"learning_rate": 4.8001465917419985e-05,
|
| 295 |
+
"loss": 1.2665,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.04104167684565398,
|
| 300 |
+
"grad_norm": 0.92928147315979,
|
| 301 |
+
"learning_rate": 4.795260200342048e-05,
|
| 302 |
+
"loss": 1.2627,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.04201885962769336,
|
| 307 |
+
"grad_norm": 1.0651229619979858,
|
| 308 |
+
"learning_rate": 4.790373808942096e-05,
|
| 309 |
+
"loss": 1.2623,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.04299604240973274,
|
| 314 |
+
"grad_norm": 0.9612557888031006,
|
| 315 |
+
"learning_rate": 4.7854874175421456e-05,
|
| 316 |
+
"loss": 1.2482,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.043973225191772124,
|
| 321 |
+
"grad_norm": 1.0120874643325806,
|
| 322 |
+
"learning_rate": 4.7806010261421944e-05,
|
| 323 |
+
"loss": 1.2589,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.0449504079738115,
|
| 328 |
+
"grad_norm": 0.6250020861625671,
|
| 329 |
+
"learning_rate": 4.775714634742243e-05,
|
| 330 |
+
"loss": 1.2499,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.04592759075585088,
|
| 335 |
+
"grad_norm": 0.2850038707256317,
|
| 336 |
+
"learning_rate": 4.770828243342292e-05,
|
| 337 |
+
"loss": 1.2446,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.046904773537890264,
|
| 342 |
+
"grad_norm": 1.2032625675201416,
|
| 343 |
+
"learning_rate": 4.765941851942341e-05,
|
| 344 |
+
"loss": 1.2238,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.04788195631992964,
|
| 349 |
+
"grad_norm": 0.42024949193000793,
|
| 350 |
+
"learning_rate": 4.7610554605423896e-05,
|
| 351 |
+
"loss": 1.2255,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.048859139101969026,
|
| 356 |
+
"grad_norm": 0.7451406121253967,
|
| 357 |
+
"learning_rate": 4.756169069142439e-05,
|
| 358 |
+
"loss": 1.2071,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.049836321884008404,
|
| 363 |
+
"grad_norm": 0.8735096454620361,
|
| 364 |
+
"learning_rate": 4.751282677742487e-05,
|
| 365 |
+
"loss": 1.2126,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.05081350466604778,
|
| 370 |
+
"grad_norm": 0.73675137758255,
|
| 371 |
+
"learning_rate": 4.746396286342537e-05,
|
| 372 |
+
"loss": 1.2036,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.051790687448087167,
|
| 377 |
+
"grad_norm": 0.6540606617927551,
|
| 378 |
+
"learning_rate": 4.741509894942585e-05,
|
| 379 |
+
"loss": 1.1825,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.052767870230126544,
|
| 384 |
+
"grad_norm": 0.825066864490509,
|
| 385 |
+
"learning_rate": 4.7366235035426336e-05,
|
| 386 |
+
"loss": 1.1655,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.05374505301216593,
|
| 391 |
+
"grad_norm": 1.6421219110488892,
|
| 392 |
+
"learning_rate": 4.731737112142683e-05,
|
| 393 |
+
"loss": 1.1716,
|
| 394 |
+
"step": 5500
|
| 395 |
+
}
|
| 396 |
+
],
|
| 397 |
+
"logging_steps": 100,
|
| 398 |
+
"max_steps": 102335,
|
| 399 |
+
"num_input_tokens_seen": 0,
|
| 400 |
+
"num_train_epochs": 1,
|
| 401 |
+
"save_steps": 500,
|
| 402 |
+
"stateful_callbacks": {
|
| 403 |
+
"TrainerControl": {
|
| 404 |
+
"args": {
|
| 405 |
+
"should_epoch_stop": false,
|
| 406 |
+
"should_evaluate": false,
|
| 407 |
+
"should_log": false,
|
| 408 |
+
"should_save": true,
|
| 409 |
+
"should_training_stop": false
|
| 410 |
+
},
|
| 411 |
+
"attributes": {}
|
| 412 |
+
}
|
| 413 |
+
},
|
| 414 |
+
"total_flos": 6.898607743565824e+16,
|
| 415 |
+
"train_batch_size": 12,
|
| 416 |
+
"trial_name": null,
|
| 417 |
+
"trial_params": null
|
| 418 |
+
}
|
checkpoint-5500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:278b8ae4ca6b721ea0f2d20061c63ec30cd0ca06fdf7af3d153eaddb11a3fd6b
|
| 3 |
+
size 5176
|