Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-v4-v4_cleanup_code

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
82f21f1
·
verified ·
1 Parent(s): 6a3ea6c

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 sha256:79fe90cb7da18146d6d9227e57a677f0db76aa81beb9d064e69b5a9edbe0ddeb
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": 0.09811484499300901,
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": 0.0004821368304324767,
14
+ "grad_norm": 84.9173377406223,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 1.5154205322265626,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0009642736608649534,
21
+ "grad_norm": 57.560206229549436,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 0.9048759460449218,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.0014464104912974302,
28
+ "grad_norm": 241.93706017750705,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.7844394207000732,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.001928547321729907,
35
+ "grad_norm": 635.4867043743025,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.777440071105957,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.0024106841521623836,
42
+ "grad_norm": 16.70216416699672,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.7813506126403809,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.0028928209825948605,
49
+ "grad_norm": 104.65219520703347,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.7643991470336914,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0033749578130273373,
56
+ "grad_norm": 15.464766046614894,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.7663109779357911,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.003857094643459814,
63
+ "grad_norm": 11.140861449353931,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.7194852352142334,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.004339231473892291,
70
+ "grad_norm": 29.54062487849426,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.7484091758728028,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.004821368304324767,
77
+ "grad_norm": 623.2770461417654,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.7809892177581788,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.0053035051347572445,
84
+ "grad_norm": 12.453456400945432,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7879758358001709,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.005785641965189721,
91
+ "grad_norm": 15.266718921768598,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7485491752624511,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.006267778795622197,
98
+ "grad_norm": 7.765117181857264,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.7564555168151855,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.006749915626054675,
105
+ "grad_norm": 12.596599218999913,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.7486157894134522,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.007232052456487151,
112
+ "grad_norm": 4.964019715699265,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.6995878219604492,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.007714189286919628,
119
+ "grad_norm": 5.34953585002889,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.7818653106689453,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.008196326117352104,
126
+ "grad_norm": 13.677128436034314,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.7409422874450684,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.008678462947784581,
133
+ "grad_norm": 11.253318513799217,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.7045968532562256,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.009160599778217059,
140
+ "grad_norm": 3.7993902470656638,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.7415530204772949,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.009642736608649534,
147
+ "grad_norm": 9.932013471995777,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.7458738327026367,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.010124873439082012,
154
+ "grad_norm": 4.001842841007721,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.7419017791748047,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.010607010269514489,
161
+ "grad_norm": 5.8081003010500405,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.727513599395752,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.011089147099946964,
168
+ "grad_norm": 24.899690781901665,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.7438565731048584,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.011571283930379442,
175
+ "grad_norm": 9.431012313223512,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.7601432800292969,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.012053420760811919,
182
+ "grad_norm": 9.9069654154098,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.7237838268280029,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.012535557591244395,
189
+ "grad_norm": 6.065436761140742,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.7469722747802734,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.013017694421676872,
196
+ "grad_norm": 47.854432068582106,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.7887792587280273,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.01349983125210935,
203
+ "grad_norm": 7.739139876114809,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.8068546295166016,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.013981968082541825,
210
+ "grad_norm": 6.596407728810894,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.7717693328857422,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.014464104912974302,
217
+ "grad_norm": 2.821767608514682,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.762939453125,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.01494624174340678,
224
+ "grad_norm": 12.92389489442801,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.7905154705047608,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.015428378573839255,
231
+ "grad_norm": 4.45893615790758,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.8191452026367188,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.01591051540427173,
238
+ "grad_norm": 2.719882613980345,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.7672964096069336,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.016392652234704208,
245
+ "grad_norm": 9.433039678311145,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.7420453548431396,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.016874789065136685,
252
+ "grad_norm": 4.012690253492805,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.7563004493713379,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.017356925895569163,
259
+ "grad_norm": 11.624966196988954,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.7373766899108887,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.01783906272600164,
266
+ "grad_norm": 2.7344222258766258,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.7051375389099122,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.018321199556434117,
273
+ "grad_norm": 4.539182866411412,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.7277828216552734,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.01880333638686659,
280
+ "grad_norm": 4.87684609900965,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.7499900817871094,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.01928547321729907,
287
+ "grad_norm": 2.305121309654471,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.7218913078308106,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.019767610047731546,
294
+ "grad_norm": 2.637730882882061,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.7323745250701904,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.020249746878164023,
301
+ "grad_norm": 2.9314732121215092,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.7070964336395263,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.0207318837085965,
308
+ "grad_norm": 4.051033736245248,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.7526719570159912,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.021214020539028978,
315
+ "grad_norm": 2.737987761268551,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.7400177001953125,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.02169615736946145,
322
+ "grad_norm": 2.002819925510059,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.7192610740661621,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.02217829419989393,
329
+ "grad_norm": 3.7426192848860755,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.7366347312927246,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.022660431030326406,
336
+ "grad_norm": 1.1926626716478357,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.7299184322357177,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.023142567860758884,
343
+ "grad_norm": 3.367060431673956,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.7061259269714355,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.02362470469119136,
350
+ "grad_norm": 1.0695774993776388,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7048762321472168,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.024106841521623838,
357
+ "grad_norm": 1.9484852417237744,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.7045268058776856,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.024588978352056312,
364
+ "grad_norm": 1.7161008623305876,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.7366765022277832,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.02507111518248879,
371
+ "grad_norm": 2.7468813266571335,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.7303312301635743,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.025553252012921267,
378
+ "grad_norm": 8.640342170047546,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.76505126953125,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.026035388843353744,
385
+ "grad_norm": 14.403207471659123,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.7315347194671631,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.02651752567378622,
392
+ "grad_norm": 4.9187469514726025,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.7313735961914063,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.0269996625042187,
399
+ "grad_norm": 21.09742751886116,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.7265950679779053,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.027481799334651173,
406
+ "grad_norm": 9.120234530490755,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.752630615234375,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.02796393616508365,
413
+ "grad_norm": 39.217881534027256,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.7522613048553467,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.028446072995516127,
420
+ "grad_norm": 15.15469395513839,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.7306804656982422,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.028928209825948605,
427
+ "grad_norm": 2.229860272443176,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.7174323081970215,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.029410346656381082,
434
+ "grad_norm": 28.87034725965148,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.7655416488647461,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.02989248348681356,
441
+ "grad_norm": 2.85440650717933,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7155353546142578,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.030374620317246033,
448
+ "grad_norm": 4.824295325677904,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.7334116935729981,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.03085675714767851,
455
+ "grad_norm": 6.533365855916318,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.7016949653625488,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.03133889397811099,
462
+ "grad_norm": 16.877036489930155,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7247958660125733,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.03182103080854346,
469
+ "grad_norm": 329.2790796078927,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.7393420219421387,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.03230316763897594,
476
+ "grad_norm": 3.076950012733951,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.762078857421875,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.032785304469408416,
483
+ "grad_norm": 4.030194347977115,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.7703542232513427,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.0332674412998409,
490
+ "grad_norm": 1.813412222853007,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.7255609035491943,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.03374957813027337,
497
+ "grad_norm": 3.3819806682901143,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.7121047973632812,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.03423171496070585,
504
+ "grad_norm": 2.3872601241148304,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.706441068649292,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.034713851791138325,
511
+ "grad_norm": 3.5426833842605157,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.7012114524841309,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.0351959886215708,
518
+ "grad_norm": 4.564258584135792,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.723219633102417,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.03567812545200328,
525
+ "grad_norm": 3.4441825412487086,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.6977805137634278,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.036160262282435754,
532
+ "grad_norm": 3.8546226650995714,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.7184113979339599,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.036642399112868235,
539
+ "grad_norm": 16.213966783350067,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.7428165435791015,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.03712453594330071,
546
+ "grad_norm": 2.0741210758941913,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.7358497619628906,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.03760667277373318,
553
+ "grad_norm": 2.5202301046172417,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.7103074550628662,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.03808880960416566,
560
+ "grad_norm": 4.106110045969379,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.713107967376709,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.03857094643459814,
567
+ "grad_norm": 3.088072216074999,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7298454284667969,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.03905308326503062,
574
+ "grad_norm": 2.171527717000864,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.7251861095428467,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.03953522009546309,
581
+ "grad_norm": 2.0872404223379406,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.7178170680999756,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.04001735692589557,
588
+ "grad_norm": 2.085406019030795,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.6847468376159668,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.040499493756328046,
595
+ "grad_norm": 6.295654776698936,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.7038663387298584,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.04098163058676052,
602
+ "grad_norm": 2.454918336179734,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.7551318645477295,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.041463767417193,
609
+ "grad_norm": 2.5362213989256044,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.7268366813659668,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.041945904247625475,
616
+ "grad_norm": 2.0309978712081254,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.725460433959961,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.042428041078057956,
623
+ "grad_norm": 2.726392751612776,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.7183332443237305,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.04291017790849043,
630
+ "grad_norm": 2.2408894596393836,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.7466308116912842,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.0433923147389229,
637
+ "grad_norm": 8.581913026001844,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.67788667678833,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.043874451569355384,
644
+ "grad_norm": 14.329361671424175,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.6972944259643554,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.04435658839978786,
651
+ "grad_norm": 3.9088010842539327,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.699748182296753,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.04483872523022034,
658
+ "grad_norm": 2.221525537997574,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.7268280982971191,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.04532086206065281,
665
+ "grad_norm": 2.273319630076157,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.7030668258666992,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.04580299889108529,
672
+ "grad_norm": 2.8492569998023027,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.7284522533416748,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.04628513572151777,
679
+ "grad_norm": 3.6240853365585473,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.6911632061004639,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.04676727255195024,
686
+ "grad_norm": 4.035089048393918,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.7237444877624511,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.04724940938238272,
693
+ "grad_norm": 1.7484668255186857,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.7165475368499756,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.047731546212815196,
700
+ "grad_norm": 3.6565231516157466,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.7023948669433594,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.048213683043247677,
707
+ "grad_norm": 8.83591745574243,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.714229679107666,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.04869581987368015,
714
+ "grad_norm": 1.5477438520033022,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.6836585998535156,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.049177956704112624,
721
+ "grad_norm": 1.7108650020940415,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.6963685512542724,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.049660093534545105,
728
+ "grad_norm": 3.8025775202775516,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.7289498329162598,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.05014223036497758,
735
+ "grad_norm": 1.6346781298258524,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.7262353897094727,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.05062436719541006,
742
+ "grad_norm": 3.386679850071031,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.6808836936950684,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.051106504025842533,
749
+ "grad_norm": 2.3953301062602854,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.7089850425720214,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.051588640856275014,
756
+ "grad_norm": 2.5139615513318763,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.7181517601013183,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.05207077768670749,
763
+ "grad_norm": 2.8534620949787053,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.6958971977233886,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.05255291451713996,
770
+ "grad_norm": 1.9564028862046592,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.7039243698120117,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.05303505134757244,
777
+ "grad_norm": 1.557897571396239,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.7114754676818847,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.05351718817800492,
784
+ "grad_norm": 2.930498345415913,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.7031359672546387,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.0539993250084374,
791
+ "grad_norm": 1.4602419678677463,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.7077861785888672,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.05448146183886987,
798
+ "grad_norm": 1.7319226933333385,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.7047904968261719,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.054963598669302345,
805
+ "grad_norm": 2.520611756871299,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.7063097953796387,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.055445735499734826,
812
+ "grad_norm": 3.214756094588082,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.6928376197814942,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.0559278723301673,
819
+ "grad_norm": 2.1913033139614546,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.7101710319519043,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.05641000916059978,
826
+ "grad_norm": 5.747581126680708,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.7223403453826904,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.056892145991032254,
833
+ "grad_norm": 1.478178600197385,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.7083209991455078,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.05737428282146473,
840
+ "grad_norm": 3.1153048058866166,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.7255118370056153,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.05785641965189721,
847
+ "grad_norm": 6.6985613434510975,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.726648473739624,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.05833855648232968,
854
+ "grad_norm": 5.377966618368178,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.7175424575805665,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.058820693312762164,
861
+ "grad_norm": 3.838988528518626,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.7453009605407714,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.05930283014319464,
868
+ "grad_norm": 2.6228691909450834,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.7184647560119629,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.05978496697362712,
875
+ "grad_norm": 2.6612776295017615,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.7309318542480469,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.06026710380405959,
882
+ "grad_norm": 1.8845329907929567,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.725009822845459,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.060749240634492066,
889
+ "grad_norm": 10.28021708941531,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.7213550567626953,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.06123137746492455,
896
+ "grad_norm": 1.7751708481039612,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.6989108562469483,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.06171351429535702,
903
+ "grad_norm": 2.058733549275171,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.6917882442474366,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.0621956511257895,
910
+ "grad_norm": 5.491929938751169,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.6999297142028809,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.06267778795622198,
917
+ "grad_norm": 6.688351091581926,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.7235936164855957,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.06315992478665446,
924
+ "grad_norm": 4.899979463669896,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.686549711227417,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.06364206161708692,
931
+ "grad_norm": 2.952614374772642,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.7102313041687012,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.0641241984475194,
938
+ "grad_norm": 10.566954823585363,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.7124320983886718,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.06460633527795188,
945
+ "grad_norm": 8.873763489301671,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.7242676734924316,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.06508847210838437,
952
+ "grad_norm": 3.170822217707811,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.709017276763916,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.06557060893881683,
959
+ "grad_norm": 4.133385298753556,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.7052340507507324,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.06605274576924931,
966
+ "grad_norm": 2.3980143582956575,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.7251494884490967,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.0665348825996818,
973
+ "grad_norm": 1.8149128248978232,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.6977263450622558,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.06701701943011426,
980
+ "grad_norm": 3.216048638536296,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.7132225036621094,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.06749915626054674,
987
+ "grad_norm": 2.759251898652736,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.6781229972839355,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.06798129309097922,
994
+ "grad_norm": 2.1615844816494523,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.6893566131591797,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.0684634299214117,
1001
+ "grad_norm": 1.269324197852931,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.7151976585388183,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.06894556675184417,
1008
+ "grad_norm": 2.475140950996921,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.7208847045898438,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.06942770358227665,
1015
+ "grad_norm": 1.8637431609519701,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.6920453548431397,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.06990984041270913,
1022
+ "grad_norm": 4.388822465185116,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.7050337791442871,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.0703919772431416,
1029
+ "grad_norm": 1.7686798808274666,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.6877009391784668,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.07087411407357408,
1036
+ "grad_norm": 2.1589166054135314,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.6840052604675293,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.07135625090400656,
1043
+ "grad_norm": 3.4176787553601677,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.7044760704040527,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.07183838773443903,
1050
+ "grad_norm": 5.058389900187091,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.7252198219299316,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.07232052456487151,
1057
+ "grad_norm": 3.4902103862821328,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.7132100582122802,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.07280266139530399,
1064
+ "grad_norm": 6.018184255339297,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.70932936668396,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.07328479822573647,
1071
+ "grad_norm": 3.5144758180231968,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.7115926742553711,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.07376693505616894,
1078
+ "grad_norm": 6.3900031796310985,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.7138595581054688,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.07424907188660142,
1085
+ "grad_norm": 1.830421739776605,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.7291291236877442,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.0747312087170339,
1092
+ "grad_norm": 4.409354344559001,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.7407273292541504,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.07521334554746636,
1099
+ "grad_norm": 3.4473627534438513,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.7394618511199951,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.07569548237789885,
1106
+ "grad_norm": 2.6382177594854754,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.6957945823669434,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.07617761920833133,
1113
+ "grad_norm": 3.999056912619744,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.6612532615661622,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.07665975603876381,
1120
+ "grad_norm": 3.7247803590530237,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.7023423194885254,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.07714189286919627,
1127
+ "grad_norm": 1.8337543544271353,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.7308431148529053,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.07762402969962875,
1134
+ "grad_norm": 4.832144300014454,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.7019494533538818,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.07810616653006124,
1141
+ "grad_norm": 1.2557192725218083,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.7272754669189453,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.0785883033604937,
1148
+ "grad_norm": 1.259086888724499,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.7088966369628906,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.07907044019092618,
1155
+ "grad_norm": 1.4497993722505094,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.7055980682373046,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.07955257702135866,
1162
+ "grad_norm": 1.9410198139904098,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.7067210674285889,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.08003471385179114,
1169
+ "grad_norm": 1.8227117633516707,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.6959988594055175,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.08051685068222361,
1176
+ "grad_norm": 8.864878739103949,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.7231966018676758,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.08099898751265609,
1183
+ "grad_norm": 5.91600501179535,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.6998160362243653,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.08148112434308857,
1190
+ "grad_norm": 2.7073467229023582,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.7311582088470459,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.08196326117352104,
1197
+ "grad_norm": 2.6882475083061217,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.7005412101745605,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.08244539800395352,
1204
+ "grad_norm": 1.3644811855655996,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.7040036678314209,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.082927534834386,
1211
+ "grad_norm": 1.707401344240697,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.7216098308563232,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.08340967166481847,
1218
+ "grad_norm": 2.2473714034468366,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.7012883186340332,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.08389180849525095,
1225
+ "grad_norm": 3.8729069145758053,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.7171290397644043,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.08437394532568343,
1232
+ "grad_norm": 2.2548793711676427,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.6937798976898193,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.08485608215611591,
1239
+ "grad_norm": 3.363398202084773,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.7099102973937989,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.08533821898654838,
1246
+ "grad_norm": 1.9683460905836172,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.7375545501708984,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.08582035581698086,
1253
+ "grad_norm": 2.7563303203565104,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.693428373336792,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.08630249264741334,
1260
+ "grad_norm": 5.488470914545985,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.6957249641418457,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.0867846294778458,
1267
+ "grad_norm": 2.09336496768559,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.6892104148864746,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.08726676630827829,
1274
+ "grad_norm": 1.5894667449541344,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.6863278388977051,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.08774890313871077,
1281
+ "grad_norm": 2.351872874716584,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.6925557136535645,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.08823103996914325,
1288
+ "grad_norm": 5.709628555644903,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.7161927223205566,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.08871317679957572,
1295
+ "grad_norm": 2.3637395099075658,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.6833971977233887,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.0891953136300082,
1302
+ "grad_norm": 3.0782660457463638,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.6962731838226318,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.08967745046044068,
1309
+ "grad_norm": 3.0074368386337524,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.6978533744812012,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.09015958729087314,
1316
+ "grad_norm": 2.887814419604526,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.7000442504882812,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.09064172412130563,
1323
+ "grad_norm": 3.6280575625203513,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.7079222679138184,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.0911238609517381,
1330
+ "grad_norm": 1.3557511666014126,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.6967864036560059,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.09160599778217059,
1337
+ "grad_norm": 1.403629260906539,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.7130854606628418,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.09208813461260305,
1344
+ "grad_norm": 1.8739359959248085,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.6871781826019288,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.09257027144303553,
1351
+ "grad_norm": 2.824655544812341,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.6981027126312256,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.09305240827346802,
1358
+ "grad_norm": 1.4828172680841498,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.6929860591888428,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.09353454510390048,
1365
+ "grad_norm": 7.2089191719005425,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.7134679794311524,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.09401668193433296,
1372
+ "grad_norm": 2.1889125436919574,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.7014190196990967,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.09449881876476544,
1379
+ "grad_norm": 6.312627812926516,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.7476691722869873,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.09498095559519791,
1386
+ "grad_norm": 1.2052636178685785,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.7070323944091796,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.09546309242563039,
1393
+ "grad_norm": 1.1645089780675213,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.6702715873718261,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.09594522925606287,
1400
+ "grad_norm": 3.612031603825751,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.6985786437988282,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.09642736608649535,
1407
+ "grad_norm": 1.2061047374120728,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.6996285915374756,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.09690950291692782,
1414
+ "grad_norm": 1.319276109312496,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.693095588684082,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.0973916397473603,
1421
+ "grad_norm": 1.358389710976587,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.7233096122741699,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.09787377657779278,
1428
+ "grad_norm": 2.29102697168753,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.699147367477417,
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:e04372c0377560216662710081ce91385f58bcf95862bebd29a7102741e0347e
3
+ size 7313