Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g1_cleanup_code

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
1279f55
·
verified ·
1 Parent(s): 9c6b726

Upload folder using huggingface_hub

Browse files
Files changed (24) hide show
  1. .gitattributes +1 -0
  2. 200M/chat_template.jinja +1 -0
  3. 200M/config.json +170 -0
  4. 200M/generation_config.json +17 -0
  5. 200M/latest +1 -0
  6. 200M/model.safetensors +3 -0
  7. 200M/processor_config.json +62 -0
  8. 200M/rng_state_0.pth +3 -0
  9. 200M/rng_state_1.pth +3 -0
  10. 200M/rng_state_10.pth +3 -0
  11. 200M/rng_state_11.pth +3 -0
  12. 200M/rng_state_2.pth +3 -0
  13. 200M/rng_state_3.pth +3 -0
  14. 200M/rng_state_4.pth +3 -0
  15. 200M/rng_state_5.pth +3 -0
  16. 200M/rng_state_6.pth +3 -0
  17. 200M/rng_state_7.pth +3 -0
  18. 200M/rng_state_8.pth +3 -0
  19. 200M/rng_state_9.pth +3 -0
  20. 200M/scheduler.pt +3 -0
  21. 200M/tokenizer.json +3 -0
  22. 200M/tokenizer_config.json +56 -0
  23. 200M/trainer_state.json +1455 -0
  24. 200M/training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ 200M/tokenizer.json filter=lfs diff=lfs merge=lfs -text
200M/chat_template.jinja ADDED
@@ -0,0 +1 @@
 
 
1
+ {% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content }}{% endif %}{% endfor %}
200M/config.json ADDED
@@ -0,0 +1,170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma4UnifiedForConditionalGeneration"
4
+ ],
5
+ "audio_config": {
6
+ "_name_or_path": "",
7
+ "architectures": null,
8
+ "audio_embed_dim": 640,
9
+ "chunk_size_feed_forward": 0,
10
+ "dtype": "bfloat16",
11
+ "id2label": {
12
+ "0": "LABEL_0",
13
+ "1": "LABEL_1"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "is_encoder_decoder": false,
17
+ "label2id": {
18
+ "LABEL_0": 0,
19
+ "LABEL_1": 1
20
+ },
21
+ "model_type": "gemma4_unified_audio",
22
+ "output_attentions": false,
23
+ "output_hidden_states": false,
24
+ "problem_type": null,
25
+ "return_dict": true,
26
+ "rms_norm_eps": 1e-06
27
+ },
28
+ "audio_token_id": 258881,
29
+ "boa_token_id": 256000,
30
+ "boi_token_id": 255999,
31
+ "bos_token_id": 2,
32
+ "dtype": "bfloat16",
33
+ "eoa_token_index": 258883,
34
+ "eoi_token_id": 258882,
35
+ "eos_token_id": 1,
36
+ "hidden_size": 3840,
37
+ "image_token_id": 258880,
38
+ "initializer_range": 0.02,
39
+ "model_type": "gemma4_unified",
40
+ "pad_token_id": 0,
41
+ "text_config": {
42
+ "attention_bias": false,
43
+ "attention_dropout": 0.0,
44
+ "attention_k_eq_v": true,
45
+ "bos_token_id": 2,
46
+ "dtype": "bfloat16",
47
+ "enable_moe_block": false,
48
+ "eos_token_id": 1,
49
+ "final_logit_softcapping": 30.0,
50
+ "global_head_dim": 512,
51
+ "head_dim": 256,
52
+ "hidden_activation": "gelu_pytorch_tanh",
53
+ "hidden_size": 3840,
54
+ "hidden_size_per_layer_input": 0,
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 15360,
57
+ "layer_types": [
58
+ "sliding_attention",
59
+ "sliding_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "full_attention",
64
+ "sliding_attention",
65
+ "sliding_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "full_attention",
70
+ "sliding_attention",
71
+ "sliding_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "full_attention",
76
+ "sliding_attention",
77
+ "sliding_attention",
78
+ "sliding_attention",
79
+ "sliding_attention",
80
+ "sliding_attention",
81
+ "full_attention",
82
+ "sliding_attention",
83
+ "sliding_attention",
84
+ "sliding_attention",
85
+ "sliding_attention",
86
+ "sliding_attention",
87
+ "full_attention",
88
+ "sliding_attention",
89
+ "sliding_attention",
90
+ "sliding_attention",
91
+ "sliding_attention",
92
+ "sliding_attention",
93
+ "full_attention",
94
+ "sliding_attention",
95
+ "sliding_attention",
96
+ "sliding_attention",
97
+ "sliding_attention",
98
+ "sliding_attention",
99
+ "full_attention",
100
+ "sliding_attention",
101
+ "sliding_attention",
102
+ "sliding_attention",
103
+ "sliding_attention",
104
+ "sliding_attention",
105
+ "full_attention"
106
+ ],
107
+ "max_position_embeddings": 262144,
108
+ "model_type": "gemma4_unified_text",
109
+ "moe_intermediate_size": null,
110
+ "num_attention_heads": 16,
111
+ "num_experts": null,
112
+ "num_global_key_value_heads": 1,
113
+ "num_hidden_layers": 48,
114
+ "num_key_value_heads": 8,
115
+ "num_kv_shared_layers": 0,
116
+ "pad_token_id": 0,
117
+ "rms_norm_eps": 1e-06,
118
+ "rope_parameters": {
119
+ "full_attention": {
120
+ "partial_rotary_factor": 0.25,
121
+ "rope_theta": 1000000.0,
122
+ "rope_type": "proportional"
123
+ },
124
+ "sliding_attention": {
125
+ "rope_theta": 10000.0,
126
+ "rope_type": "default"
127
+ }
128
+ },
129
+ "sliding_window": 1024,
130
+ "tie_word_embeddings": true,
131
+ "top_k_experts": null,
132
+ "use_bidirectional_attention": "vision",
133
+ "use_cache": false,
134
+ "use_double_wide_mlp": false,
135
+ "vocab_size": 262144,
136
+ "vocab_size_per_layer_input": 262144
137
+ },
138
+ "tie_word_embeddings": true,
139
+ "transformers_version": "5.10.0",
140
+ "use_cache": false,
141
+ "video_token_id": 258884,
142
+ "vision_config": {
143
+ "_name_or_path": "",
144
+ "architectures": null,
145
+ "chunk_size_feed_forward": 0,
146
+ "dtype": "bfloat16",
147
+ "id2label": {
148
+ "0": "LABEL_0",
149
+ "1": "LABEL_1"
150
+ },
151
+ "initializer_range": 0.02,
152
+ "is_encoder_decoder": false,
153
+ "label2id": {
154
+ "LABEL_0": 0,
155
+ "LABEL_1": 1
156
+ },
157
+ "mm_embed_dim": 3840,
158
+ "mm_posemb_size": 1120,
159
+ "model_type": "gemma4_unified_vision",
160
+ "num_soft_tokens": 280,
161
+ "output_attentions": false,
162
+ "output_hidden_states": false,
163
+ "output_proj_dims": 3840,
164
+ "patch_size": 16,
165
+ "pooling_kernel_size": 3,
166
+ "problem_type": null,
167
+ "return_dict": true,
168
+ "rms_norm_eps": 1e-06
169
+ }
170
+ }
200M/generation_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 1,
6
+ 1
7
+ ],
8
+ "pad_token_id": 0,
9
+ "suppress_tokens": [
10
+ 258883,
11
+ 258882
12
+ ],
13
+ "temperature": 1.0,
14
+ "top_k": 64,
15
+ "top_p": 0.95,
16
+ "transformers_version": "5.10.0"
17
+ }
200M/latest ADDED
@@ -0,0 +1 @@
 
 
1
+ global_step2035
200M/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb0033e8f412ebcb6f7d29a07a35a634360f2b965c0c276f30d0550fda95ed20
3
+ size 23919549408
200M/processor_config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_ms_per_token": 40,
3
+ "audio_seq_length": 750,
4
+ "feature_extractor": {
5
+ "audio_samples_per_token": 640,
6
+ "feature_extractor_type": "Gemma4UnifiedAudioFeatureExtractor",
7
+ "feature_size": 640,
8
+ "padding_side": "right",
9
+ "padding_value": 0.0,
10
+ "return_attention_mask": true,
11
+ "sampling_rate": 16000
12
+ },
13
+ "image_processor": {
14
+ "do_convert_rgb": true,
15
+ "do_normalize": false,
16
+ "do_rescale": true,
17
+ "do_resize": true,
18
+ "image_mean": [
19
+ 0.0,
20
+ 0.0,
21
+ 0.0
22
+ ],
23
+ "image_processor_type": "Gemma4UnifiedImageProcessor",
24
+ "image_std": [
25
+ 1.0,
26
+ 1.0,
27
+ 1.0
28
+ ],
29
+ "max_soft_tokens": 280,
30
+ "patch_size": 16,
31
+ "pooling_kernel_size": 3,
32
+ "resample": 3,
33
+ "rescale_factor": 0.00392156862745098
34
+ },
35
+ "image_seq_length": 280,
36
+ "processor_class": "Gemma4UnifiedProcessor",
37
+ "video_processor": {
38
+ "do_convert_rgb": true,
39
+ "do_normalize": true,
40
+ "do_rescale": true,
41
+ "do_resize": true,
42
+ "do_sample_frames": true,
43
+ "image_mean": [
44
+ 0.0,
45
+ 0.0,
46
+ 0.0
47
+ ],
48
+ "image_std": [
49
+ 1.0,
50
+ 1.0,
51
+ 1.0
52
+ ],
53
+ "max_soft_tokens": 70,
54
+ "num_frames": 32,
55
+ "patch_size": 16,
56
+ "pooling_kernel_size": 3,
57
+ "resample": 3,
58
+ "rescale_factor": 0.00392156862745098,
59
+ "return_metadata": false,
60
+ "video_processor_type": "Gemma4UnifiedVideoProcessor"
61
+ }
62
+ }
200M/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d7dcf54a8dbd1fff243640ddc034149fbe7c90ed70b8a95a6bb093e989a7577
3
+ size 14469
200M/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2a2b0b9d330140afd742a57df45106276d425d38a78e53b0fd43dba9559f8d4
3
+ size 14469
200M/rng_state_10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ce222f8f11fd62fd6cb82264500aefbc9e7ef16907ed3e4aa3a8a99c21d95d3
3
+ size 14476
200M/rng_state_11.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d07144bfc4ed675dda70ab66923d17d6f9233560b2db09689cbe1f262c590c88
3
+ size 14476
200M/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5db634f6ed5151d3c898fbe3c365f950516c404bed926de16b253b26e63c7cf
3
+ size 14469
200M/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad774eade44525098fff2d5910e3d16565fca4c2af29ed0bc4d9315a8ac9607a
3
+ size 14469
200M/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebdee1b37be2238764bcfeed9dad268e938248152b0fba60eaf89adda99eac0f
3
+ size 14469
200M/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6046157bd3d69a7cf6ee508b7359b30cb71488a11b11b07a614041b06813c553
3
+ size 14469
200M/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebc9300f101a324d0361f01aec1d57e26b57a703b4e536f4fe982c8ba57a9144
3
+ size 14469
200M/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:357df7b609d6ffd7f22b998281f7856a711ee83505e5ed2835959554b0185df7
3
+ size 14469
200M/rng_state_8.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fb33667d752a94ade6841f62b8b8fcc22256b4712bd568b141bc690d4a42e11
3
+ size 14469
200M/rng_state_9.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:354a7a38e97ba11b8ad29bf4d00d74f92295b32124cff0ef46baa862aa9b0f2f
3
+ size 14469
200M/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea084231345496123a1e2e3e6e6377891838cfabbf1b42517ab7575e5966916e
3
+ size 1465
200M/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12bac982b793c44b03d52a250a9f0d0b666813da566b910c24a6da0695fd11e6
3
+ size 32170070
200M/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_token": "<|audio|>",
3
+ "backend": "tokenizers",
4
+ "boa_token": "<|audio>",
5
+ "boi_token": "<|image>",
6
+ "bos_token": "<bos>",
7
+ "eoa_token": "<audio|>",
8
+ "eoc_token": "<channel|>",
9
+ "eoi_token": "<image|>",
10
+ "eos_token": "<eos>",
11
+ "eot_token": "<turn|>",
12
+ "escape_token": "<|\"|>",
13
+ "etc_token": "<tool_call|>",
14
+ "etd_token": "<tool|>",
15
+ "etr_token": "<tool_response|>",
16
+ "extra_special_tokens": [
17
+ "<|video|>"
18
+ ],
19
+ "image_token": "<|image|>",
20
+ "is_local": true,
21
+ "local_files_only": true,
22
+ "mask_token": "<mask>",
23
+ "model_max_length": 1000000000000000019884624838656,
24
+ "model_specific_special_tokens": {
25
+ "audio_token": "<|audio|>",
26
+ "boa_token": "<|audio>",
27
+ "boi_token": "<|image>",
28
+ "eoa_token": "<audio|>",
29
+ "eoc_token": "<channel|>",
30
+ "eoi_token": "<image|>",
31
+ "eot_token": "<turn|>",
32
+ "escape_token": "<|\"|>",
33
+ "etc_token": "<tool_call|>",
34
+ "etd_token": "<tool|>",
35
+ "etr_token": "<tool_response|>",
36
+ "image_token": "<|image|>",
37
+ "soc_token": "<|channel>",
38
+ "sot_token": "<|turn>",
39
+ "stc_token": "<|tool_call>",
40
+ "std_token": "<|tool>",
41
+ "str_token": "<|tool_response>",
42
+ "think_token": "<|think|>"
43
+ },
44
+ "pad_token": "<pad>",
45
+ "padding_side": "right",
46
+ "processor_class": "Gemma4UnifiedProcessor",
47
+ "soc_token": "<|channel>",
48
+ "sot_token": "<|turn>",
49
+ "split_special_tokens": false,
50
+ "stc_token": "<|tool_call>",
51
+ "std_token": "<|tool>",
52
+ "str_token": "<|tool_response>",
53
+ "think_token": "<|think|>",
54
+ "tokenizer_class": "GemmaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
200M/trainer_state.json ADDED
@@ -0,0 +1,1455 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.12679127725856698,
6
+ "eval_steps": 500,
7
+ "global_step": 2035,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0006230529595015577,
14
+ "grad_norm": 63.15375442917419,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 1.372981071472168,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0012461059190031153,
21
+ "grad_norm": 63.39470944322245,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 0.8310403823852539,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.001869158878504673,
28
+ "grad_norm": 11.96310343076372,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.7284352779388428,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0024922118380062306,
35
+ "grad_norm": 159.35287956464992,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.6918130874633789,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.003115264797507788,
42
+ "grad_norm": 140.43690439070613,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.6914069652557373,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.003738317757009346,
49
+ "grad_norm": 20.04034887568955,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.6684707641601563,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.004361370716510903,
56
+ "grad_norm": 146.62158267929294,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.6722880363464355,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.004984423676012461,
63
+ "grad_norm": 48.33263649592231,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.7312785625457764,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.005607476635514018,
70
+ "grad_norm": 96.97131898292652,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.7157273292541504,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.006230529595015576,
77
+ "grad_norm": 95.5113244321525,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.7347418785095214,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.006853582554517134,
84
+ "grad_norm": 17.213242453649954,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7522170066833496,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.007476635514018692,
91
+ "grad_norm": 31.300913667908624,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7286031246185303,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.00809968847352025,
98
+ "grad_norm": 69.2417895023605,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.6891061305999756,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.008722741433021807,
105
+ "grad_norm": 20.324536486476994,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 1.3122027397155762,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.009345794392523364,
112
+ "grad_norm": 4.902224214231,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.7708694458007812,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.009968847352024923,
119
+ "grad_norm": 14.386566577478696,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.7179630279541016,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.01059190031152648,
126
+ "grad_norm": 7.998488720594685,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.684850025177002,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.011214953271028037,
133
+ "grad_norm": 10.20788593785828,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.7212257385253906,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.011838006230529595,
140
+ "grad_norm": 15.093739760534373,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.7274382591247559,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.012461059190031152,
147
+ "grad_norm": 7.624987181272481,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.6997462272644043,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.013084112149532711,
154
+ "grad_norm": 38.20889799071737,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.8538092613220215,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.013707165109034268,
161
+ "grad_norm": 5.610917778335814,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.7119334697723388,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.014330218068535825,
168
+ "grad_norm": 7.071002934278071,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.6945069313049317,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.014953271028037384,
175
+ "grad_norm": 1.9779972566834714,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.6839393615722656,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.01557632398753894,
182
+ "grad_norm": 4.542653465163337,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.6552486419677734,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.0161993769470405,
189
+ "grad_norm": 4.1326478566474645,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.7250523567199707,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.016822429906542057,
196
+ "grad_norm": 2.71061736504502,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.6365773677825928,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.017445482866043614,
203
+ "grad_norm": 2.4835202986671043,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.6632792472839355,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.01806853582554517,
210
+ "grad_norm": 22.09800173446232,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.663812780380249,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.018691588785046728,
217
+ "grad_norm": 15.489771776735218,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.650740385055542,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.019314641744548288,
224
+ "grad_norm": 1.8136045854235476,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.6493139266967773,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.019937694704049845,
231
+ "grad_norm": 1.4242810637346537,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.6703683376312256,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.020560747663551402,
238
+ "grad_norm": 2.01645414429854,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.6692399501800537,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.02118380062305296,
245
+ "grad_norm": 1.3748624267038327,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.6543716430664063,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.021806853582554516,
252
+ "grad_norm": 3.7632290163066595,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.6510223388671875,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.022429906542056073,
259
+ "grad_norm": 3.069062402265582,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.6401685237884521,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.023052959501557634,
266
+ "grad_norm": 3.9242151311515756,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.650885534286499,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.02367601246105919,
273
+ "grad_norm": 2.5645994240054146,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.6438897132873536,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.024299065420560748,
280
+ "grad_norm": 3.3264331329128587,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.6713788032531738,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.024922118380062305,
287
+ "grad_norm": 5.952680558627565,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.695007610321045,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.02554517133956386,
294
+ "grad_norm": 7.2640539794724965,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.6785341262817383,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.026168224299065422,
301
+ "grad_norm": 12.065849294819978,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.6561185836791992,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.02679127725856698,
308
+ "grad_norm": 2.9882502235892647,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.6439939498901367,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.027414330218068536,
315
+ "grad_norm": 2.919046299610538,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.6663602352142334,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.028037383177570093,
322
+ "grad_norm": 4.189627201982762,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.6693576812744141,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.02866043613707165,
329
+ "grad_norm": 7.890660674041811,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.6684383392333985,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.029283489096573207,
336
+ "grad_norm": 82.91478815091692,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7290635585784913,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.029906542056074768,
343
+ "grad_norm": 3.333359402201608,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.6626537322998047,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.030529595015576325,
350
+ "grad_norm": 2.4351561694941313,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.6529546260833741,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.03115264797507788,
357
+ "grad_norm": 3.5168056840413233,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.6785277843475341,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.03177570093457944,
364
+ "grad_norm": 64.52415387460657,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.6317290306091309,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.032398753894081,
371
+ "grad_norm": 1.9567345855965799,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.6504751205444336,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.03302180685358255,
378
+ "grad_norm": 2.4483916233070815,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.6714927673339843,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.03364485981308411,
385
+ "grad_norm": 6.84493838576302,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.6702317714691162,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.03426791277258567,
392
+ "grad_norm": 1.6302467458157637,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.6749137878417969,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.03489096573208723,
399
+ "grad_norm": 1.3830723147413793,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.6726428985595703,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.03551401869158879,
406
+ "grad_norm": 1.816565044561178,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.6633450984954834,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.03613707165109034,
413
+ "grad_norm": 1.4154803760531598,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.6670496940612793,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.0367601246105919,
420
+ "grad_norm": 1.441388864695392,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.66356520652771,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.037383177570093455,
427
+ "grad_norm": 5.297501644827849,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.6568848609924316,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.038006230529595016,
434
+ "grad_norm": 1.3268959282636765,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.6364224433898926,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.038629283489096576,
441
+ "grad_norm": 3.1357427218379725,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.6608361721038818,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.03925233644859813,
448
+ "grad_norm": 3.2870153380544496,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.7057211875915528,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.03987538940809969,
455
+ "grad_norm": 2.197431012725625,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.6897117614746093,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.040498442367601244,
462
+ "grad_norm": 8.502549125517568,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.6920805931091308,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.041121495327102804,
469
+ "grad_norm": 2.462320802295259,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.7063193321228027,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.041744548286604365,
476
+ "grad_norm": 5.683201827218238,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.6920748710632324,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.04236760124610592,
483
+ "grad_norm": 3.027115327083748,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.6580988883972168,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.04299065420560748,
490
+ "grad_norm": 9.069631138625143,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.6772819519042969,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.04361370716510903,
497
+ "grad_norm": 3.5222303907627026,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.6700127601623536,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.04423676012461059,
504
+ "grad_norm": 2.539625946046831,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.6825683116912842,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.044859813084112146,
511
+ "grad_norm": 7.530973621569606,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.6668391227722168,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.04548286604361371,
518
+ "grad_norm": 3.10826200653366,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.6900657176971435,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.04610591900311527,
525
+ "grad_norm": 3.021206626575616,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.666568374633789,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.04672897196261682,
532
+ "grad_norm": 3.4063444253551793,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.6786983489990235,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.04735202492211838,
539
+ "grad_norm": 5.674285755423112,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.6631364822387695,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.047975077881619935,
546
+ "grad_norm": 5.447979039303706,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.6851268768310547,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.048598130841121495,
553
+ "grad_norm": 2.0548995784325275,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.6662421703338623,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.049221183800623056,
560
+ "grad_norm": 1.2276649895732483,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.6671768188476562,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.04984423676012461,
567
+ "grad_norm": 1.5987199277283939,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.6489299774169922,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.05046728971962617,
574
+ "grad_norm": 22.9031846877308,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.6675164699554443,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.05109034267912772,
581
+ "grad_norm": 2.9034364102338985,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.6826793670654296,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.051713395638629284,
588
+ "grad_norm": 1.8432534502921918,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.6763291835784913,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.052336448598130844,
595
+ "grad_norm": 2.0417799918298596,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.6497424125671387,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.0529595015576324,
602
+ "grad_norm": 22.32540564160995,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.6354857444763183,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.05358255451713396,
609
+ "grad_norm": 6.229762616389806,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.657370662689209,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.05420560747663551,
616
+ "grad_norm": 3.23998905516123,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.6524142265319824,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.05482866043613707,
623
+ "grad_norm": 2.8784021094456604,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.653789472579956,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.05545171339563863,
630
+ "grad_norm": 1.40857663500608,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.6703225612640381,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.056074766355140186,
637
+ "grad_norm": 5.080917667561178,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.6439192771911622,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.05669781931464175,
644
+ "grad_norm": 4.5764665849587445,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.6634146690368652,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.0573208722741433,
651
+ "grad_norm": 3.2037710370417023,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.6824374198913574,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.05794392523364486,
658
+ "grad_norm": 6.053033149131505,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.6423258781433105,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.058566978193146414,
665
+ "grad_norm": 13.436486964861611,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.6521054744720459,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.059190031152647975,
672
+ "grad_norm": 1.4820242941537698,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.6490562438964844,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.059813084112149535,
679
+ "grad_norm": 3.970984796336633,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.6425233364105225,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.06043613707165109,
686
+ "grad_norm": 2.606688835969579,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.6780929088592529,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.06105919003115265,
693
+ "grad_norm": 1.6441407042964533,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.6381833553314209,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.0616822429906542,
700
+ "grad_norm": 1.9338014114509088,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.671638298034668,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.06230529595015576,
707
+ "grad_norm": 2.39864419114088,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.659054183959961,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.06292834890965732,
714
+ "grad_norm": 4.683446585267705,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.6786028861999511,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.06355140186915888,
721
+ "grad_norm": 2.719894383949491,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.6658808708190918,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.06417445482866044,
728
+ "grad_norm": 1.1588652494157246,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.6696397304534912,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.064797507788162,
735
+ "grad_norm": 1.0813640353054965,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.6676293849945069,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.06542056074766354,
742
+ "grad_norm": 1.2271064002015928,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.6389595985412597,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.0660436137071651,
749
+ "grad_norm": 2.8431458204110074,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.6555405616760254,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.06666666666666667,
756
+ "grad_norm": 3.2188265816780692,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.6418774604797364,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.06728971962616823,
763
+ "grad_norm": 9.465426552560904,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.6477942943572998,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.06791277258566979,
770
+ "grad_norm": 1.1426314291875073,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.6486814022064209,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.06853582554517133,
777
+ "grad_norm": 1.3614517863008955,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.6399056911468506,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.0691588785046729,
784
+ "grad_norm": 1.616958777603617,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.625498104095459,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.06978193146417445,
791
+ "grad_norm": 1.700623675066532,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.620833158493042,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.07040498442367601,
798
+ "grad_norm": 2.5588029176909948,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.6351099967956543,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.07102803738317758,
805
+ "grad_norm": 6.556254605023658,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.6419850826263428,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.07165109034267912,
812
+ "grad_norm": 2.4751544762969946,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.6722092628479004,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.07227414330218068,
819
+ "grad_norm": 1.5727047413222857,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.6558211326599122,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.07289719626168224,
826
+ "grad_norm": 9.438322770095112,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.6519765853881836,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.0735202492211838,
833
+ "grad_norm": 2.839029932122009,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.6472910881042481,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.07414330218068536,
840
+ "grad_norm": 1.3804888984788395,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.6604276180267334,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.07476635514018691,
847
+ "grad_norm": 1.0934274555239194,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.6465075969696045,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.07538940809968847,
854
+ "grad_norm": 1.8841069578655136,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.6400872230529785,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.07601246105919003,
861
+ "grad_norm": 16.12011257195063,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.655942440032959,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.07663551401869159,
868
+ "grad_norm": 1.9392045965213456,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.6418064117431641,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.07725856697819315,
875
+ "grad_norm": 1.2879388752480592,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.6534547328948974,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.0778816199376947,
882
+ "grad_norm": 1.3491269021694952,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.6325421333312988,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.07850467289719626,
889
+ "grad_norm": 1.9538179100083624,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.6555644989013671,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.07912772585669782,
896
+ "grad_norm": 1.686205721804821,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.6385934352874756,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.07975077881619938,
903
+ "grad_norm": 1.4326411502688534,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.6342977046966553,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.08037383177570094,
910
+ "grad_norm": 4.860024808022459,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.6609707832336426,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.08099688473520249,
917
+ "grad_norm": 1.2371686758942715,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.6476799964904785,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.08161993769470405,
924
+ "grad_norm": 1.1597203515801051,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.6473023891448975,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.08224299065420561,
931
+ "grad_norm": 1.8273978345410087,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.6425920009613038,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.08286604361370717,
938
+ "grad_norm": 1.6088638146203655,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.6325952053070069,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.08348909657320873,
945
+ "grad_norm": 1.0395526162178266,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.6280010223388672,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.08411214953271028,
952
+ "grad_norm": 1.1776111039430426,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.6420888900756836,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.08473520249221184,
959
+ "grad_norm": 1.144362156326936,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.6447340965270996,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.0853582554517134,
966
+ "grad_norm": 1.2899619747320272,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.6319891929626464,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.08598130841121496,
973
+ "grad_norm": 1.1503180888416502,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.6263866424560547,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.08660436137071652,
980
+ "grad_norm": 0.9492449615285058,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.6156892776489258,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.08722741433021806,
987
+ "grad_norm": 0.9840501517202644,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6201627731323243,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.08785046728971962,
994
+ "grad_norm": 0.9663965412569069,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.618841552734375,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.08847352024922119,
1001
+ "grad_norm": 1.2024904063187882,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.6406625270843506,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.08909657320872275,
1008
+ "grad_norm": 0.9573647223711624,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.6550570487976074,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.08971962616822429,
1015
+ "grad_norm": 1.0307197954823704,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.6338574886322021,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.09034267912772585,
1022
+ "grad_norm": 2.305281565846579,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.6363538265228271,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.09096573208722741,
1029
+ "grad_norm": 1.4667417153142859,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.6235406875610352,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.09158878504672897,
1036
+ "grad_norm": 2.9584530551772437,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.6403841018676758,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.09221183800623053,
1043
+ "grad_norm": 1.112253956223209,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.6414080619812011,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.09283489096573208,
1050
+ "grad_norm": 1.197396557670252,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.6136775970458984,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.09345794392523364,
1057
+ "grad_norm": 0.930725260856421,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.6281547546386719,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.0940809968847352,
1064
+ "grad_norm": 2.246282067394849,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.6232186317443847,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.09470404984423676,
1071
+ "grad_norm": 1.811695916729897,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.6236707210540772,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.09532710280373832,
1078
+ "grad_norm": 1.8453018493553404,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.6396072387695313,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.09595015576323987,
1085
+ "grad_norm": 1.4810091845213549,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.6346172332763672,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.09657320872274143,
1092
+ "grad_norm": 1.1558342150707985,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.6340350151062012,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.09719626168224299,
1099
+ "grad_norm": 1.1277930144534245,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.6178334712982178,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.09781931464174455,
1106
+ "grad_norm": 1.6697265802855836,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6273056030273437,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.09844236760124611,
1113
+ "grad_norm": 1.1996090977492178,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.6388472557067871,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.09906542056074766,
1120
+ "grad_norm": 4.193851985080715,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.6388599395751953,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.09968847352024922,
1127
+ "grad_norm": 1.3073390628519994,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.6429434776306152,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.10031152647975078,
1134
+ "grad_norm": 1.5006350300856774,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.6440896034240723,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.10093457943925234,
1141
+ "grad_norm": 1.2275982378982346,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.6285616874694824,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.1015576323987539,
1148
+ "grad_norm": 1.1626753929589895,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.6467322826385498,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.10218068535825545,
1155
+ "grad_norm": 1.8546526164062445,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.6216934204101563,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.102803738317757,
1162
+ "grad_norm": 4.204227006721846,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.6037077903747559,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.10342679127725857,
1169
+ "grad_norm": 2.4575617089230892,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.6549998760223389,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.10404984423676013,
1176
+ "grad_norm": 0.9831225747692173,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.6362233638763428,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.10467289719626169,
1183
+ "grad_norm": 37.336116705288426,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.6452502250671387,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.10529595015576323,
1190
+ "grad_norm": 10.151830562748597,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.6224545478820801,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.1059190031152648,
1197
+ "grad_norm": 1.0474870590574359,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.6610437393188476,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.10654205607476636,
1204
+ "grad_norm": 3.461782785767983,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.62818922996521,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.10716510903426792,
1211
+ "grad_norm": 1.0436258381858183,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.6486543178558349,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.10778816199376948,
1218
+ "grad_norm": 3.870674655703194,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.6329143524169922,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.10841121495327102,
1225
+ "grad_norm": 0.8873679451298351,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.6266915798187256,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.10903426791277258,
1232
+ "grad_norm": 1.934393771485732,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.6206174373626709,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.10965732087227414,
1239
+ "grad_norm": 0.9788497030191609,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.617664098739624,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.1102803738317757,
1246
+ "grad_norm": 1.3294647839474172,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.6125317573547363,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.11090342679127727,
1253
+ "grad_norm": 0.9114337186956374,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.6136180877685546,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.11152647975077881,
1260
+ "grad_norm": 0.9168910070389206,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.6254988670349121,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.11214953271028037,
1267
+ "grad_norm": 1.2163942422286602,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.6339495658874512,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.11277258566978193,
1274
+ "grad_norm": 1.0031026506964655,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.6163856506347656,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.1133956386292835,
1281
+ "grad_norm": 1.4487390671485545,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.6172023296356202,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.11401869158878504,
1288
+ "grad_norm": 0.971335799491775,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.603542423248291,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.1146417445482866,
1295
+ "grad_norm": 1.0645808741308516,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.6314529418945313,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.11526479750778816,
1302
+ "grad_norm": 1.1564622773995759,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.6358022212982177,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.11588785046728972,
1309
+ "grad_norm": 6.250954738770854,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.6266725063323975,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.11651090342679128,
1316
+ "grad_norm": 2.485950620772887,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.6347331047058106,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.11713395638629283,
1323
+ "grad_norm": 1.0491315047430319,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.6353681087493896,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.11775700934579439,
1330
+ "grad_norm": 0.9344118553673864,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.6243095397949219,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.11838006230529595,
1337
+ "grad_norm": 1.166751507604473,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.624453353881836,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.11900311526479751,
1344
+ "grad_norm": 1.1609173008743447,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.6156741142272949,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.11962616822429907,
1351
+ "grad_norm": 1.2650824767917728,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.6225091457366944,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.12024922118380062,
1358
+ "grad_norm": 1.323039674561904,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.6126754760742188,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.12087227414330218,
1365
+ "grad_norm": 10.576149324131405,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.6002795219421386,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.12149532710280374,
1372
+ "grad_norm": 1.1408899621175108,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.6268309593200684,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.1221183800623053,
1379
+ "grad_norm": 1.04557318312074,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.6239619255065918,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.12274143302180686,
1386
+ "grad_norm": 1.2159263913891547,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.6297917842864991,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.1233644859813084,
1393
+ "grad_norm": 2.784941022644952,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6152359008789062,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.12398753894080997,
1400
+ "grad_norm": 0.9318872113204291,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.6123355388641357,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.12461059190031153,
1407
+ "grad_norm": 1.6289119810248258,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.6148693084716796,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.1252336448598131,
1414
+ "grad_norm": 6.458481915206926,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.6158293724060059,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.12585669781931463,
1421
+ "grad_norm": 1.9163812666195066,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.6355224132537842,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.1264797507788162,
1428
+ "grad_norm": 5.359098164590273,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.6285907745361328,
1431
+ "step": 2030
1432
+ }
1433
+ ],
1434
+ "logging_steps": 10,
1435
+ "max_steps": 5935,
1436
+ "num_input_tokens_seen": 0,
1437
+ "num_train_epochs": 1,
1438
+ "save_steps": 2035,
1439
+ "stateful_callbacks": {
1440
+ "TrainerControl": {
1441
+ "args": {
1442
+ "should_epoch_stop": false,
1443
+ "should_evaluate": false,
1444
+ "should_log": false,
1445
+ "should_save": true,
1446
+ "should_training_stop": false
1447
+ },
1448
+ "attributes": {}
1449
+ }
1450
+ },
1451
+ "total_flos": 963613493821440.0,
1452
+ "train_batch_size": 1,
1453
+ "trial_name": null,
1454
+ "trial_params": null
1455
+ }
200M/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aff51fc471427e3e0a6d8c88d0bfeec583ddb6b2423d3b49640f42c3d156eb52
3
+ size 7313