Training in progress, step 6500, checkpoint
Browse files- checkpoint-6500/config.json +30 -0
- checkpoint-6500/generation_config.json +6 -0
- checkpoint-6500/model.safetensors +3 -0
- checkpoint-6500/optimizer.pt +3 -0
- checkpoint-6500/rng_state.pth +3 -0
- checkpoint-6500/scheduler.pt +3 -0
- checkpoint-6500/special_tokens_map.json +24 -0
- checkpoint-6500/tokenizer.json +96 -0
- checkpoint-6500/tokenizer_config.json +38 -0
- checkpoint-6500/trainer_state.json +488 -0
- checkpoint-6500/training_args.bin +3 -0
checkpoint-6500/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "EleutherAI/pythia-14m",
|
| 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": 128,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 512,
|
| 16 |
+
"layer_norm_eps": 1e-05,
|
| 17 |
+
"max_position_embeddings": 2048,
|
| 18 |
+
"model_type": "gpt_neox",
|
| 19 |
+
"num_attention_heads": 4,
|
| 20 |
+
"num_hidden_layers": 6,
|
| 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-6500/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-6500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74001a99aabee8985901bc143a81bf56b65bdf929f80c5e6e02014d8f4e8c67d
|
| 3 |
+
size 4776024
|
checkpoint-6500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f9372244df1e8bc2c58eed3647b134d20dd29cfbed0d0584ddcfddbbf78dd8c
|
| 3 |
+
size 9599034
|
checkpoint-6500/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-6500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8733f9d674cf5a86001b86e40941e9f953e3e78750c6e6290b239a72f460a8a1
|
| 3 |
+
size 1064
|
checkpoint-6500/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-6500/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 |
+
"<|endoftext|>": 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-6500/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-6500/trainer_state.json
ADDED
|
@@ -0,0 +1,488 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.2540650406504065,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 6500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.003908692933083177,
|
| 13 |
+
"grad_norm": 0.6370952129364014,
|
| 14 |
+
"learning_rate": 4.98240400406663e-05,
|
| 15 |
+
"loss": 1.4791,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.007817385866166354,
|
| 20 |
+
"grad_norm": 0.27393946051597595,
|
| 21 |
+
"learning_rate": 4.9628528974739973e-05,
|
| 22 |
+
"loss": 1.3743,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01172607879924953,
|
| 27 |
+
"grad_norm": 0.364713191986084,
|
| 28 |
+
"learning_rate": 4.943301790881364e-05,
|
| 29 |
+
"loss": 1.3525,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.015634771732332707,
|
| 34 |
+
"grad_norm": 0.33364734053611755,
|
| 35 |
+
"learning_rate": 4.923750684288731e-05,
|
| 36 |
+
"loss": 1.3457,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.019543464665415886,
|
| 41 |
+
"grad_norm": 0.3672705292701721,
|
| 42 |
+
"learning_rate": 4.904199577696098e-05,
|
| 43 |
+
"loss": 1.3375,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02345215759849906,
|
| 48 |
+
"grad_norm": 0.16828787326812744,
|
| 49 |
+
"learning_rate": 4.884648471103465e-05,
|
| 50 |
+
"loss": 1.3348,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02736085053158224,
|
| 55 |
+
"grad_norm": 0.24634529650211334,
|
| 56 |
+
"learning_rate": 4.8650973645108314e-05,
|
| 57 |
+
"loss": 1.3343,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.031269543464665414,
|
| 62 |
+
"grad_norm": 0.5399248003959656,
|
| 63 |
+
"learning_rate": 4.845546257918198e-05,
|
| 64 |
+
"loss": 1.3287,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03517823639774859,
|
| 69 |
+
"grad_norm": 0.5724157691001892,
|
| 70 |
+
"learning_rate": 4.825995151325565e-05,
|
| 71 |
+
"loss": 1.3236,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.03908692933083177,
|
| 76 |
+
"grad_norm": 0.5199714303016663,
|
| 77 |
+
"learning_rate": 4.8064440447329324e-05,
|
| 78 |
+
"loss": 1.3189,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04299562226391495,
|
| 83 |
+
"grad_norm": 0.30912989377975464,
|
| 84 |
+
"learning_rate": 4.786892938140299e-05,
|
| 85 |
+
"loss": 1.315,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.04690431519699812,
|
| 90 |
+
"grad_norm": 0.37046098709106445,
|
| 91 |
+
"learning_rate": 4.767341831547666e-05,
|
| 92 |
+
"loss": 1.3151,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0508130081300813,
|
| 97 |
+
"grad_norm": 0.26445892453193665,
|
| 98 |
+
"learning_rate": 4.747790724955033e-05,
|
| 99 |
+
"loss": 1.3118,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.05472170106316448,
|
| 104 |
+
"grad_norm": 0.3207770586013794,
|
| 105 |
+
"learning_rate": 4.728239618362399e-05,
|
| 106 |
+
"loss": 1.3087,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.05863039399624766,
|
| 111 |
+
"grad_norm": 0.4711158573627472,
|
| 112 |
+
"learning_rate": 4.7086885117697664e-05,
|
| 113 |
+
"loss": 1.31,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.06253908692933083,
|
| 118 |
+
"grad_norm": 0.524340808391571,
|
| 119 |
+
"learning_rate": 4.689137405177133e-05,
|
| 120 |
+
"loss": 1.3062,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.06644777986241401,
|
| 125 |
+
"grad_norm": 0.3912469744682312,
|
| 126 |
+
"learning_rate": 4.6695862985845e-05,
|
| 127 |
+
"loss": 1.3078,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07035647279549719,
|
| 132 |
+
"grad_norm": 0.3625667691230774,
|
| 133 |
+
"learning_rate": 4.6500351919918674e-05,
|
| 134 |
+
"loss": 1.3039,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.07426516572858036,
|
| 139 |
+
"grad_norm": 0.3897973299026489,
|
| 140 |
+
"learning_rate": 4.630484085399234e-05,
|
| 141 |
+
"loss": 1.3027,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.07817385866166354,
|
| 146 |
+
"grad_norm": 0.30837342143058777,
|
| 147 |
+
"learning_rate": 4.6109329788066005e-05,
|
| 148 |
+
"loss": 1.3055,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.08208255159474671,
|
| 153 |
+
"grad_norm": 0.3017924129962921,
|
| 154 |
+
"learning_rate": 4.591381872213967e-05,
|
| 155 |
+
"loss": 1.3021,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.0859912445278299,
|
| 160 |
+
"grad_norm": 0.5234686136245728,
|
| 161 |
+
"learning_rate": 4.571830765621334e-05,
|
| 162 |
+
"loss": 1.3054,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.08989993746091307,
|
| 167 |
+
"grad_norm": 0.3180347681045532,
|
| 168 |
+
"learning_rate": 4.5522796590287015e-05,
|
| 169 |
+
"loss": 1.3016,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.09380863039399624,
|
| 174 |
+
"grad_norm": 0.22097885608673096,
|
| 175 |
+
"learning_rate": 4.532728552436068e-05,
|
| 176 |
+
"loss": 1.3014,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.09771732332707943,
|
| 181 |
+
"grad_norm": 0.3503361642360687,
|
| 182 |
+
"learning_rate": 4.513177445843435e-05,
|
| 183 |
+
"loss": 1.3004,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.1016260162601626,
|
| 188 |
+
"grad_norm": 0.3880802392959595,
|
| 189 |
+
"learning_rate": 4.493626339250802e-05,
|
| 190 |
+
"loss": 1.2992,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.10553470919324578,
|
| 195 |
+
"grad_norm": 0.3632693886756897,
|
| 196 |
+
"learning_rate": 4.474075232658168e-05,
|
| 197 |
+
"loss": 1.3009,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.10944340212632896,
|
| 202 |
+
"grad_norm": 0.4512663781642914,
|
| 203 |
+
"learning_rate": 4.4545241260655355e-05,
|
| 204 |
+
"loss": 1.2999,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.11335209505941213,
|
| 209 |
+
"grad_norm": 0.3041514456272125,
|
| 210 |
+
"learning_rate": 4.434973019472902e-05,
|
| 211 |
+
"loss": 1.3007,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.11726078799249531,
|
| 216 |
+
"grad_norm": 0.34578946232795715,
|
| 217 |
+
"learning_rate": 4.415421912880269e-05,
|
| 218 |
+
"loss": 1.2985,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.12116948092557848,
|
| 223 |
+
"grad_norm": 0.2616026997566223,
|
| 224 |
+
"learning_rate": 4.3958708062876365e-05,
|
| 225 |
+
"loss": 1.2984,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.12507817385866166,
|
| 230 |
+
"grad_norm": 0.253749817609787,
|
| 231 |
+
"learning_rate": 4.376319699695003e-05,
|
| 232 |
+
"loss": 1.2963,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.12898686679174484,
|
| 237 |
+
"grad_norm": 0.24605727195739746,
|
| 238 |
+
"learning_rate": 4.35676859310237e-05,
|
| 239 |
+
"loss": 1.3005,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.13289555972482803,
|
| 244 |
+
"grad_norm": 0.31458768248558044,
|
| 245 |
+
"learning_rate": 4.337217486509736e-05,
|
| 246 |
+
"loss": 1.3008,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.13680425265791119,
|
| 251 |
+
"grad_norm": 0.23786158859729767,
|
| 252 |
+
"learning_rate": 4.3176663799171034e-05,
|
| 253 |
+
"loss": 1.2975,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.14071294559099437,
|
| 258 |
+
"grad_norm": 0.3280108869075775,
|
| 259 |
+
"learning_rate": 4.2981152733244706e-05,
|
| 260 |
+
"loss": 1.2987,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.14462163852407756,
|
| 265 |
+
"grad_norm": 0.25433337688446045,
|
| 266 |
+
"learning_rate": 4.278564166731837e-05,
|
| 267 |
+
"loss": 1.2972,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.1485303314571607,
|
| 272 |
+
"grad_norm": 0.2792419493198395,
|
| 273 |
+
"learning_rate": 4.2590130601392044e-05,
|
| 274 |
+
"loss": 1.2944,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.1524390243902439,
|
| 279 |
+
"grad_norm": 0.34164726734161377,
|
| 280 |
+
"learning_rate": 4.239461953546571e-05,
|
| 281 |
+
"loss": 1.298,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.15634771732332708,
|
| 286 |
+
"grad_norm": 0.2886170744895935,
|
| 287 |
+
"learning_rate": 4.219910846953938e-05,
|
| 288 |
+
"loss": 1.2948,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.16025641025641027,
|
| 293 |
+
"grad_norm": 0.3007793426513672,
|
| 294 |
+
"learning_rate": 4.2003597403613047e-05,
|
| 295 |
+
"loss": 1.2954,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.16416510318949343,
|
| 300 |
+
"grad_norm": 0.32882916927337646,
|
| 301 |
+
"learning_rate": 4.180808633768671e-05,
|
| 302 |
+
"loss": 1.2942,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.1680737961225766,
|
| 307 |
+
"grad_norm": 0.2950049936771393,
|
| 308 |
+
"learning_rate": 4.1612575271760384e-05,
|
| 309 |
+
"loss": 1.2938,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.1719824890556598,
|
| 314 |
+
"grad_norm": 0.7425180673599243,
|
| 315 |
+
"learning_rate": 4.141706420583405e-05,
|
| 316 |
+
"loss": 1.2926,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.17589118198874296,
|
| 321 |
+
"grad_norm": 0.26747408509254456,
|
| 322 |
+
"learning_rate": 4.122155313990772e-05,
|
| 323 |
+
"loss": 1.2933,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.17979987492182614,
|
| 328 |
+
"grad_norm": 0.3664441704750061,
|
| 329 |
+
"learning_rate": 4.1026042073981394e-05,
|
| 330 |
+
"loss": 1.2916,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.18370856785490933,
|
| 335 |
+
"grad_norm": 0.32603704929351807,
|
| 336 |
+
"learning_rate": 4.083053100805506e-05,
|
| 337 |
+
"loss": 1.2934,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.18761726078799248,
|
| 342 |
+
"grad_norm": 0.4145171046257019,
|
| 343 |
+
"learning_rate": 4.0635019942128725e-05,
|
| 344 |
+
"loss": 1.2922,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.19152595372107567,
|
| 349 |
+
"grad_norm": 0.42517679929733276,
|
| 350 |
+
"learning_rate": 4.04395088762024e-05,
|
| 351 |
+
"loss": 1.293,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.19543464665415886,
|
| 356 |
+
"grad_norm": 0.48030540347099304,
|
| 357 |
+
"learning_rate": 4.024399781027606e-05,
|
| 358 |
+
"loss": 1.2906,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.199343339587242,
|
| 363 |
+
"grad_norm": 0.4900824725627899,
|
| 364 |
+
"learning_rate": 4.0048486744349735e-05,
|
| 365 |
+
"loss": 1.293,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.2032520325203252,
|
| 370 |
+
"grad_norm": 0.400856614112854,
|
| 371 |
+
"learning_rate": 3.98529756784234e-05,
|
| 372 |
+
"loss": 1.2903,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.20716072545340838,
|
| 377 |
+
"grad_norm": 0.3311282694339752,
|
| 378 |
+
"learning_rate": 3.965746461249707e-05,
|
| 379 |
+
"loss": 1.2917,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.21106941838649157,
|
| 384 |
+
"grad_norm": 0.5248217582702637,
|
| 385 |
+
"learning_rate": 3.946195354657074e-05,
|
| 386 |
+
"loss": 1.2913,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.21497811131957473,
|
| 391 |
+
"grad_norm": 0.6463488936424255,
|
| 392 |
+
"learning_rate": 3.92664424806444e-05,
|
| 393 |
+
"loss": 1.2899,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.2188868042526579,
|
| 398 |
+
"grad_norm": 0.3202870786190033,
|
| 399 |
+
"learning_rate": 3.9070931414718075e-05,
|
| 400 |
+
"loss": 1.2866,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.2227954971857411,
|
| 405 |
+
"grad_norm": 0.5758823752403259,
|
| 406 |
+
"learning_rate": 3.887542034879174e-05,
|
| 407 |
+
"loss": 1.2907,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.22670419011882426,
|
| 412 |
+
"grad_norm": 0.4410454034805298,
|
| 413 |
+
"learning_rate": 3.867990928286541e-05,
|
| 414 |
+
"loss": 1.2885,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.23061288305190744,
|
| 419 |
+
"grad_norm": 0.3980095088481903,
|
| 420 |
+
"learning_rate": 3.8484398216939085e-05,
|
| 421 |
+
"loss": 1.2895,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.23452157598499063,
|
| 426 |
+
"grad_norm": 0.3670996427536011,
|
| 427 |
+
"learning_rate": 3.828888715101275e-05,
|
| 428 |
+
"loss": 1.2888,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.23843026891807378,
|
| 433 |
+
"grad_norm": 0.302168071269989,
|
| 434 |
+
"learning_rate": 3.8093376085086416e-05,
|
| 435 |
+
"loss": 1.2874,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.24233896185115697,
|
| 440 |
+
"grad_norm": 0.4420079290866852,
|
| 441 |
+
"learning_rate": 3.789786501916008e-05,
|
| 442 |
+
"loss": 1.2849,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.24624765478424016,
|
| 447 |
+
"grad_norm": 0.4248519539833069,
|
| 448 |
+
"learning_rate": 3.770235395323375e-05,
|
| 449 |
+
"loss": 1.2853,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.2501563477173233,
|
| 454 |
+
"grad_norm": 0.5483678579330444,
|
| 455 |
+
"learning_rate": 3.7506842887307426e-05,
|
| 456 |
+
"loss": 1.2883,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.2540650406504065,
|
| 461 |
+
"grad_norm": 0.7262341380119324,
|
| 462 |
+
"learning_rate": 3.731133182138109e-05,
|
| 463 |
+
"loss": 1.2893,
|
| 464 |
+
"step": 6500
|
| 465 |
+
}
|
| 466 |
+
],
|
| 467 |
+
"logging_steps": 100,
|
| 468 |
+
"max_steps": 25584,
|
| 469 |
+
"num_input_tokens_seen": 0,
|
| 470 |
+
"num_train_epochs": 1,
|
| 471 |
+
"save_steps": 500,
|
| 472 |
+
"stateful_callbacks": {
|
| 473 |
+
"TrainerControl": {
|
| 474 |
+
"args": {
|
| 475 |
+
"should_epoch_stop": false,
|
| 476 |
+
"should_evaluate": false,
|
| 477 |
+
"should_log": false,
|
| 478 |
+
"should_save": true,
|
| 479 |
+
"should_training_stop": false
|
| 480 |
+
},
|
| 481 |
+
"attributes": {}
|
| 482 |
+
}
|
| 483 |
+
},
|
| 484 |
+
"total_flos": 4565785116672000.0,
|
| 485 |
+
"train_batch_size": 48,
|
| 486 |
+
"trial_name": null,
|
| 487 |
+
"trial_params": null
|
| 488 |
+
}
|
checkpoint-6500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90ec914d144549ecac00847be5b8c2e93ddca3eea1598f7317b62d35eddf48d3
|
| 3 |
+
size 5176
|