Hugging Face's logo

SynthCode-Project
/
rv2-gemma12b-g2_5_plan_newcode

Safetensors
Model card Files
xet
 Community
rqadri commited on
Commit
871ea17
·
verified ·
1 Parent(s): c3df465

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:b3799a5d8fdb289ac40d3aaf882d0b559ef193414290dcd241f8b46540f5aee6
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": 0.09234050276794628,
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": 0.00045376168436337235,
14
+ "grad_norm": 33.700575849311655,
15
+ "learning_rate": 2.910112359550562e-06,
16
+ "loss": 2.135559844970703,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.0009075233687267447,
21
+ "grad_norm": 42.61937790702005,
22
+ "learning_rate": 3.921348314606742e-06,
23
+ "loss": 1.2333361625671386,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.001361285053090117,
28
+ "grad_norm": 59.78398601810045,
29
+ "learning_rate": 4.932584269662922e-06,
30
+ "loss": 0.9913753509521485,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.0018150467374534894,
35
+ "grad_norm": 9.000856351949166,
36
+ "learning_rate": 5.943820224719102e-06,
37
+ "loss": 0.8920389175415039,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.0022688084218168616,
42
+ "grad_norm": 6.411342445022513,
43
+ "learning_rate": 6.955056179775282e-06,
44
+ "loss": 0.872259521484375,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.002722570106180234,
49
+ "grad_norm": 16.593767952488978,
50
+ "learning_rate": 7.966292134831462e-06,
51
+ "loss": 0.8293242454528809,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0031763317905436064,
56
+ "grad_norm": 28.491365409414406,
57
+ "learning_rate": 8.977528089887641e-06,
58
+ "loss": 0.8571812629699707,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.003630093474906979,
63
+ "grad_norm": 70.97016809107335,
64
+ "learning_rate": 9.988764044943822e-06,
65
+ "loss": 0.8569000244140625,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.004083855159270351,
70
+ "grad_norm": 17.22979260105877,
71
+ "learning_rate": 1.1000000000000001e-05,
72
+ "loss": 0.8691472053527832,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.004537616843633723,
77
+ "grad_norm": 7.177993416395754,
78
+ "learning_rate": 1.2011235955056182e-05,
79
+ "loss": 0.847593879699707,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.004991378527997096,
84
+ "grad_norm": 86.81590963101131,
85
+ "learning_rate": 1.3022471910112362e-05,
86
+ "loss": 0.7963851451873779,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.005445140212360468,
91
+ "grad_norm": 19.345789336537777,
92
+ "learning_rate": 1.403370786516854e-05,
93
+ "loss": 0.7768198013305664,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.005898901896723841,
98
+ "grad_norm": 31.493557815634816,
99
+ "learning_rate": 1.504494382022472e-05,
100
+ "loss": 0.8187647819519043,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.006352663581087213,
105
+ "grad_norm": 28.326685019578516,
106
+ "learning_rate": 1.60561797752809e-05,
107
+ "loss": 0.8060824394226074,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.006806425265450586,
112
+ "grad_norm": 14.750887393368782,
113
+ "learning_rate": 1.706741573033708e-05,
114
+ "loss": 0.8781991958618164,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.007260186949813958,
119
+ "grad_norm": 24.02644216435556,
120
+ "learning_rate": 1.8078651685393256e-05,
121
+ "loss": 0.8258254051208496,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.00771394863417733,
126
+ "grad_norm": 12.620566889664511,
127
+ "learning_rate": 1.908988764044944e-05,
128
+ "loss": 0.8197063446044922,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.008167710318540702,
133
+ "grad_norm": 8.507375320274459,
134
+ "learning_rate": 2e-05,
135
+ "loss": 0.829745864868164,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.008621472002904074,
140
+ "grad_norm": 8.624818937885534,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.8533475875854493,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.009075233687267446,
147
+ "grad_norm": 36.279535045904844,
148
+ "learning_rate": 2e-05,
149
+ "loss": 0.8423559188842773,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.00952899537163082,
154
+ "grad_norm": 15.3885344506724,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.847628116607666,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.009982757055994192,
161
+ "grad_norm": 6.933332738558382,
162
+ "learning_rate": 2e-05,
163
+ "loss": 0.8622509002685547,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.010436518740357564,
168
+ "grad_norm": 7.288575583436198,
169
+ "learning_rate": 2e-05,
170
+ "loss": 0.8516143798828125,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.010890280424720936,
175
+ "grad_norm": 5.432497195617162,
176
+ "learning_rate": 2e-05,
177
+ "loss": 0.8451622009277344,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.01134404210908431,
182
+ "grad_norm": 13.108630640507068,
183
+ "learning_rate": 2e-05,
184
+ "loss": 0.8770389556884766,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.011797803793447682,
189
+ "grad_norm": 13.296277537542986,
190
+ "learning_rate": 2e-05,
191
+ "loss": 0.8751070976257325,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.012251565477811054,
196
+ "grad_norm": 8.15547279712017,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.8359760284423828,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.012705327162174426,
203
+ "grad_norm": 3.8614754534197484,
204
+ "learning_rate": 2e-05,
205
+ "loss": 0.8397187232971192,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.013159088846537798,
210
+ "grad_norm": 3.928378485854922,
211
+ "learning_rate": 2e-05,
212
+ "loss": 0.8067515373229981,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.013612850530901171,
217
+ "grad_norm": 14.53890809725787,
218
+ "learning_rate": 2e-05,
219
+ "loss": 0.8569300651550293,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.014066612215264543,
224
+ "grad_norm": 2.5562493023766875,
225
+ "learning_rate": 2e-05,
226
+ "loss": 0.8449411392211914,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.014520373899627915,
231
+ "grad_norm": 2.173396163551825,
232
+ "learning_rate": 2e-05,
233
+ "loss": 0.8006858825683594,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.014974135583991287,
238
+ "grad_norm": 12.884915148556473,
239
+ "learning_rate": 2e-05,
240
+ "loss": 0.8169196128845215,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.01542789726835466,
245
+ "grad_norm": 4.133267397556853,
246
+ "learning_rate": 2e-05,
247
+ "loss": 0.8178797721862793,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.015881658952718033,
252
+ "grad_norm": 4.4839538913118,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.7991756439208985,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.016335420637081405,
259
+ "grad_norm": 5.5992645013580065,
260
+ "learning_rate": 2e-05,
261
+ "loss": 0.8095648765563965,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.016789182321444777,
266
+ "grad_norm": 2.4747669764356837,
267
+ "learning_rate": 2e-05,
268
+ "loss": 0.8225090026855468,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.01724294400580815,
273
+ "grad_norm": 2.511158852648722,
274
+ "learning_rate": 2e-05,
275
+ "loss": 0.7729404449462891,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.01769670569017152,
280
+ "grad_norm": 2.2738595224738862,
281
+ "learning_rate": 2e-05,
282
+ "loss": 0.7858805656433105,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.018150467374534893,
287
+ "grad_norm": 3.0217600696491207,
288
+ "learning_rate": 2e-05,
289
+ "loss": 0.8098352432250977,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.018604229058898268,
294
+ "grad_norm": 1.8954573295840658,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.7862910747528076,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.01905799074326164,
301
+ "grad_norm": 2.0431879466326532,
302
+ "learning_rate": 2e-05,
303
+ "loss": 0.7868980407714844,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.019511752427625012,
308
+ "grad_norm": 4.6488552474284495,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.7740491390228271,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.019965514111988384,
315
+ "grad_norm": 2.1071044234720504,
316
+ "learning_rate": 2e-05,
317
+ "loss": 0.7678849697113037,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.020419275796351756,
322
+ "grad_norm": 4.824939291325391,
323
+ "learning_rate": 2e-05,
324
+ "loss": 0.7809176445007324,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.020873037480715128,
329
+ "grad_norm": 2.3615298985348807,
330
+ "learning_rate": 2e-05,
331
+ "loss": 0.8083951950073243,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.0213267991650785,
336
+ "grad_norm": 1.3296461606887355,
337
+ "learning_rate": 2e-05,
338
+ "loss": 0.8065496444702148,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.021780560849441872,
343
+ "grad_norm": 1.8850535748077886,
344
+ "learning_rate": 2e-05,
345
+ "loss": 0.7924650192260743,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.022234322533805244,
350
+ "grad_norm": 2.8212320881753956,
351
+ "learning_rate": 2e-05,
352
+ "loss": 0.7533183097839355,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.02268808421816862,
357
+ "grad_norm": 2.1567741760252757,
358
+ "learning_rate": 2e-05,
359
+ "loss": 0.8023039817810058,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.02314184590253199,
364
+ "grad_norm": 1.913414133194118,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.7970404624938965,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.023595607586895363,
371
+ "grad_norm": 7.847343548355587,
372
+ "learning_rate": 2e-05,
373
+ "loss": 0.8151960372924805,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.024049369271258735,
378
+ "grad_norm": 1.9050240664506954,
379
+ "learning_rate": 2e-05,
380
+ "loss": 0.7894588470458984,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.024503130955622107,
385
+ "grad_norm": 1.774717889208622,
386
+ "learning_rate": 2e-05,
387
+ "loss": 0.7657998561859131,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.02495689263998548,
392
+ "grad_norm": 3.19624735773649,
393
+ "learning_rate": 2e-05,
394
+ "loss": 0.8172132492065429,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.02541065432434885,
399
+ "grad_norm": 119.89007014773632,
400
+ "learning_rate": 2e-05,
401
+ "loss": 0.8138757705688476,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.025864416008712223,
406
+ "grad_norm": 3.422241769168402,
407
+ "learning_rate": 2e-05,
408
+ "loss": 0.7874475479125976,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.026318177693075595,
413
+ "grad_norm": 30.173434363153817,
414
+ "learning_rate": 2e-05,
415
+ "loss": 0.8059184074401855,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.02677193937743897,
420
+ "grad_norm": 8.636985924843131,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.9973684310913086,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.027225701061802342,
427
+ "grad_norm": 8.465993058483805,
428
+ "learning_rate": 2e-05,
429
+ "loss": 0.8142690658569336,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.027679462746165714,
434
+ "grad_norm": 7.763176456143048,
435
+ "learning_rate": 2e-05,
436
+ "loss": 0.813444709777832,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.028133224430529086,
441
+ "grad_norm": 9.151744621497256,
442
+ "learning_rate": 2e-05,
443
+ "loss": 0.7639548301696777,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.02858698611489246,
448
+ "grad_norm": 2.0256726525570907,
449
+ "learning_rate": 2e-05,
450
+ "loss": 0.8043405532836914,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.02904074779925583,
455
+ "grad_norm": 1.98716726928425,
456
+ "learning_rate": 2e-05,
457
+ "loss": 0.7883163452148437,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.029494509483619202,
462
+ "grad_norm": 2.3604683142582186,
463
+ "learning_rate": 2e-05,
464
+ "loss": 0.7694703578948975,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.029948271167982574,
469
+ "grad_norm": 13.190838557674683,
470
+ "learning_rate": 2e-05,
471
+ "loss": 0.774112319946289,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.030402032852345946,
476
+ "grad_norm": 8.133833525009468,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.7486526489257812,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.03085579453670932,
483
+ "grad_norm": 12.970450531740925,
484
+ "learning_rate": 2e-05,
485
+ "loss": 0.8198841094970704,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.03130955622107269,
490
+ "grad_norm": 18.22753344715509,
491
+ "learning_rate": 2e-05,
492
+ "loss": 0.8183542251586914,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.031763317905436066,
497
+ "grad_norm": 5.280610592205637,
498
+ "learning_rate": 2e-05,
499
+ "loss": 0.8278017044067383,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.032217079589799434,
504
+ "grad_norm": 4.20756471706492,
505
+ "learning_rate": 2e-05,
506
+ "loss": 0.8102832794189453,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.03267084127416281,
511
+ "grad_norm": 42.72438625516298,
512
+ "learning_rate": 2e-05,
513
+ "loss": 0.8131640434265137,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.033124602958526185,
518
+ "grad_norm": 3.55536319530756,
519
+ "learning_rate": 2e-05,
520
+ "loss": 0.793606185913086,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.033578364642889554,
525
+ "grad_norm": 12.959771922355795,
526
+ "learning_rate": 2e-05,
527
+ "loss": 0.7831435203552246,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.03403212632725293,
532
+ "grad_norm": 6.469259997463987,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.7753517150878906,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.0344858880116163,
539
+ "grad_norm": 34.477603328158615,
540
+ "learning_rate": 2e-05,
541
+ "loss": 0.7536171913146973,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.03493964969597967,
546
+ "grad_norm": 2.328509918402065,
547
+ "learning_rate": 2e-05,
548
+ "loss": 0.7809432029724122,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.03539341138034304,
553
+ "grad_norm": 4.288352937474369,
554
+ "learning_rate": 2e-05,
555
+ "loss": 0.7952579498291016,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.03584717306470642,
560
+ "grad_norm": 7.639274296461497,
561
+ "learning_rate": 2e-05,
562
+ "loss": 0.8167254447937011,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.036300934749069785,
567
+ "grad_norm": 2.5903744589165703,
568
+ "learning_rate": 2e-05,
569
+ "loss": 0.7940209865570068,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.03675469643343316,
574
+ "grad_norm": 1.714046167478395,
575
+ "learning_rate": 2e-05,
576
+ "loss": 0.7802738189697266,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.037208458117796536,
581
+ "grad_norm": 2.356191303105163,
582
+ "learning_rate": 2e-05,
583
+ "loss": 0.7603988170623779,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.037662219802159905,
588
+ "grad_norm": 2.6324392358450206,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.7551675319671631,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.03811598148652328,
595
+ "grad_norm": 5.1149592386678036,
596
+ "learning_rate": 2e-05,
597
+ "loss": 0.7713553428649902,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.03856974317088665,
602
+ "grad_norm": 4.250591266250503,
603
+ "learning_rate": 2e-05,
604
+ "loss": 0.7647957801818848,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.039023504855250024,
609
+ "grad_norm": 1.484157740941218,
610
+ "learning_rate": 2e-05,
611
+ "loss": 0.7875026226043701,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.03947726653961339,
616
+ "grad_norm": 2.4537172207690814,
617
+ "learning_rate": 2e-05,
618
+ "loss": 0.7982926368713379,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.03993102822397677,
623
+ "grad_norm": 1.885761099206628,
624
+ "learning_rate": 2e-05,
625
+ "loss": 0.7709912776947021,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.04038478990834014,
630
+ "grad_norm": 3.1726728603243513,
631
+ "learning_rate": 2e-05,
632
+ "loss": 0.7637313842773438,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.04083855159270351,
637
+ "grad_norm": 1.734967497093215,
638
+ "learning_rate": 2e-05,
639
+ "loss": 0.78067626953125,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 0.04129231327706689,
644
+ "grad_norm": 3.298790323372369,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.7815995693206788,
647
+ "step": 910
648
+ },
649
+ {
650
+ "epoch": 0.041746074961430256,
651
+ "grad_norm": 1.3197318794126558,
652
+ "learning_rate": 2e-05,
653
+ "loss": 0.7714223861694336,
654
+ "step": 920
655
+ },
656
+ {
657
+ "epoch": 0.04219983664579363,
658
+ "grad_norm": 2.179324182696462,
659
+ "learning_rate": 2e-05,
660
+ "loss": 0.752001142501831,
661
+ "step": 930
662
+ },
663
+ {
664
+ "epoch": 0.042653598330157,
665
+ "grad_norm": 1.7760762886617165,
666
+ "learning_rate": 2e-05,
667
+ "loss": 0.7619918346405029,
668
+ "step": 940
669
+ },
670
+ {
671
+ "epoch": 0.043107360014520375,
672
+ "grad_norm": 5.1168406185966235,
673
+ "learning_rate": 2e-05,
674
+ "loss": 0.7750636577606201,
675
+ "step": 950
676
+ },
677
+ {
678
+ "epoch": 0.043561121698883744,
679
+ "grad_norm": 1.8080346226599653,
680
+ "learning_rate": 2e-05,
681
+ "loss": 0.7890705108642578,
682
+ "step": 960
683
+ },
684
+ {
685
+ "epoch": 0.04401488338324712,
686
+ "grad_norm": 1.5578518869917433,
687
+ "learning_rate": 2e-05,
688
+ "loss": 0.7533071041107178,
689
+ "step": 970
690
+ },
691
+ {
692
+ "epoch": 0.04446864506761049,
693
+ "grad_norm": 2.0331640286970094,
694
+ "learning_rate": 2e-05,
695
+ "loss": 0.7617393493652344,
696
+ "step": 980
697
+ },
698
+ {
699
+ "epoch": 0.04492240675197386,
700
+ "grad_norm": 1.2015052654892246,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.7665273189544678,
703
+ "step": 990
704
+ },
705
+ {
706
+ "epoch": 0.04537616843633724,
707
+ "grad_norm": 1.628143392881603,
708
+ "learning_rate": 2e-05,
709
+ "loss": 0.770779037475586,
710
+ "step": 1000
711
+ },
712
+ {
713
+ "epoch": 0.04582993012070061,
714
+ "grad_norm": 1.2039882827957658,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.7887059211730957,
717
+ "step": 1010
718
+ },
719
+ {
720
+ "epoch": 0.04628369180506398,
721
+ "grad_norm": 1.5827671273859805,
722
+ "learning_rate": 2e-05,
723
+ "loss": 0.7503167629241944,
724
+ "step": 1020
725
+ },
726
+ {
727
+ "epoch": 0.04673745348942735,
728
+ "grad_norm": 3.868510887527835,
729
+ "learning_rate": 2e-05,
730
+ "loss": 0.7796614646911622,
731
+ "step": 1030
732
+ },
733
+ {
734
+ "epoch": 0.047191215173790727,
735
+ "grad_norm": 2.9140174925180067,
736
+ "learning_rate": 2e-05,
737
+ "loss": 0.8053103446960449,
738
+ "step": 1040
739
+ },
740
+ {
741
+ "epoch": 0.047644976858154095,
742
+ "grad_norm": 7.892541947564107,
743
+ "learning_rate": 2e-05,
744
+ "loss": 0.8123416900634766,
745
+ "step": 1050
746
+ },
747
+ {
748
+ "epoch": 0.04809873854251747,
749
+ "grad_norm": 1.8993885865680917,
750
+ "learning_rate": 2e-05,
751
+ "loss": 0.7850803375244141,
752
+ "step": 1060
753
+ },
754
+ {
755
+ "epoch": 0.04855250022688084,
756
+ "grad_norm": 2.1292285710118657,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.8199052810668945,
759
+ "step": 1070
760
+ },
761
+ {
762
+ "epoch": 0.049006261911244214,
763
+ "grad_norm": 1.4765373586924606,
764
+ "learning_rate": 2e-05,
765
+ "loss": 0.7806931972503662,
766
+ "step": 1080
767
+ },
768
+ {
769
+ "epoch": 0.04946002359560759,
770
+ "grad_norm": 1.516040890469186,
771
+ "learning_rate": 2e-05,
772
+ "loss": 0.7693392753601074,
773
+ "step": 1090
774
+ },
775
+ {
776
+ "epoch": 0.04991378527997096,
777
+ "grad_norm": 2.0831299604400417,
778
+ "learning_rate": 2e-05,
779
+ "loss": 0.7863246917724609,
780
+ "step": 1100
781
+ },
782
+ {
783
+ "epoch": 0.050367546964334334,
784
+ "grad_norm": 2.195827470530011,
785
+ "learning_rate": 2e-05,
786
+ "loss": 0.7650362968444824,
787
+ "step": 1110
788
+ },
789
+ {
790
+ "epoch": 0.0508213086486977,
791
+ "grad_norm": 3.203619514755763,
792
+ "learning_rate": 2e-05,
793
+ "loss": 0.7856436729431152,
794
+ "step": 1120
795
+ },
796
+ {
797
+ "epoch": 0.05127507033306108,
798
+ "grad_norm": 3.564960215102263,
799
+ "learning_rate": 2e-05,
800
+ "loss": 0.7714583396911621,
801
+ "step": 1130
802
+ },
803
+ {
804
+ "epoch": 0.051728832017424446,
805
+ "grad_norm": 2.602524659508837,
806
+ "learning_rate": 2e-05,
807
+ "loss": 0.7595517635345459,
808
+ "step": 1140
809
+ },
810
+ {
811
+ "epoch": 0.05218259370178782,
812
+ "grad_norm": 3.0105932895678977,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.7478479385375977,
815
+ "step": 1150
816
+ },
817
+ {
818
+ "epoch": 0.05263635538615119,
819
+ "grad_norm": 1.9912075171919272,
820
+ "learning_rate": 2e-05,
821
+ "loss": 0.7397075176239014,
822
+ "step": 1160
823
+ },
824
+ {
825
+ "epoch": 0.053090117070514566,
826
+ "grad_norm": 3.0032576981429444,
827
+ "learning_rate": 2e-05,
828
+ "loss": 0.7824847221374511,
829
+ "step": 1170
830
+ },
831
+ {
832
+ "epoch": 0.05354387875487794,
833
+ "grad_norm": 1.7741279587695722,
834
+ "learning_rate": 2e-05,
835
+ "loss": 0.7740449428558349,
836
+ "step": 1180
837
+ },
838
+ {
839
+ "epoch": 0.05399764043924131,
840
+ "grad_norm": 1.9161385528021593,
841
+ "learning_rate": 2e-05,
842
+ "loss": 0.7688920021057128,
843
+ "step": 1190
844
+ },
845
+ {
846
+ "epoch": 0.054451402123604685,
847
+ "grad_norm": 1.6376294125787574,
848
+ "learning_rate": 2e-05,
849
+ "loss": 0.7435236930847168,
850
+ "step": 1200
851
+ },
852
+ {
853
+ "epoch": 0.05490516380796805,
854
+ "grad_norm": 2.655987836587518,
855
+ "learning_rate": 2e-05,
856
+ "loss": 0.7560546875,
857
+ "step": 1210
858
+ },
859
+ {
860
+ "epoch": 0.05535892549233143,
861
+ "grad_norm": 1.5087730766928222,
862
+ "learning_rate": 2e-05,
863
+ "loss": 0.7511416435241699,
864
+ "step": 1220
865
+ },
866
+ {
867
+ "epoch": 0.0558126871766948,
868
+ "grad_norm": 2.9588632994615547,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.7484208583831787,
871
+ "step": 1230
872
+ },
873
+ {
874
+ "epoch": 0.05626644886105817,
875
+ "grad_norm": 13.660542975087127,
876
+ "learning_rate": 2e-05,
877
+ "loss": 0.7785249710083008,
878
+ "step": 1240
879
+ },
880
+ {
881
+ "epoch": 0.05672021054542154,
882
+ "grad_norm": 3.189563629774851,
883
+ "learning_rate": 2e-05,
884
+ "loss": 0.7544436454772949,
885
+ "step": 1250
886
+ },
887
+ {
888
+ "epoch": 0.05717397222978492,
889
+ "grad_norm": 23.612106104880457,
890
+ "learning_rate": 2e-05,
891
+ "loss": 0.7925362586975098,
892
+ "step": 1260
893
+ },
894
+ {
895
+ "epoch": 0.05762773391414829,
896
+ "grad_norm": 3.377156755225946,
897
+ "learning_rate": 2e-05,
898
+ "loss": 0.7938989162445068,
899
+ "step": 1270
900
+ },
901
+ {
902
+ "epoch": 0.05808149559851166,
903
+ "grad_norm": 1.6453871136214189,
904
+ "learning_rate": 2e-05,
905
+ "loss": 0.7648170471191407,
906
+ "step": 1280
907
+ },
908
+ {
909
+ "epoch": 0.058535257282875036,
910
+ "grad_norm": 1.7098215646357884,
911
+ "learning_rate": 2e-05,
912
+ "loss": 0.7756949424743652,
913
+ "step": 1290
914
+ },
915
+ {
916
+ "epoch": 0.058989018967238405,
917
+ "grad_norm": 3.6636095771340043,
918
+ "learning_rate": 2e-05,
919
+ "loss": 0.8291525840759277,
920
+ "step": 1300
921
+ },
922
+ {
923
+ "epoch": 0.05944278065160178,
924
+ "grad_norm": 2.177902708692178,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.7537973403930665,
927
+ "step": 1310
928
+ },
929
+ {
930
+ "epoch": 0.05989654233596515,
931
+ "grad_norm": 2.1786512930926945,
932
+ "learning_rate": 2e-05,
933
+ "loss": 0.7512475490570069,
934
+ "step": 1320
935
+ },
936
+ {
937
+ "epoch": 0.060350304020328524,
938
+ "grad_norm": 2.3889515043153953,
939
+ "learning_rate": 2e-05,
940
+ "loss": 0.7963077068328858,
941
+ "step": 1330
942
+ },
943
+ {
944
+ "epoch": 0.06080406570469189,
945
+ "grad_norm": 6.049406950820628,
946
+ "learning_rate": 2e-05,
947
+ "loss": 0.7398816585540772,
948
+ "step": 1340
949
+ },
950
+ {
951
+ "epoch": 0.06125782738905527,
952
+ "grad_norm": 1.6910884112612445,
953
+ "learning_rate": 2e-05,
954
+ "loss": 0.7709364891052246,
955
+ "step": 1350
956
+ },
957
+ {
958
+ "epoch": 0.06171158907341864,
959
+ "grad_norm": 2.355501745859088,
960
+ "learning_rate": 2e-05,
961
+ "loss": 0.7542487621307373,
962
+ "step": 1360
963
+ },
964
+ {
965
+ "epoch": 0.06216535075778201,
966
+ "grad_norm": 2.1903603191775782,
967
+ "learning_rate": 2e-05,
968
+ "loss": 0.8148787498474122,
969
+ "step": 1370
970
+ },
971
+ {
972
+ "epoch": 0.06261911244214538,
973
+ "grad_norm": 4.583080645477956,
974
+ "learning_rate": 2e-05,
975
+ "loss": 0.7330617427825927,
976
+ "step": 1380
977
+ },
978
+ {
979
+ "epoch": 0.06307287412650876,
980
+ "grad_norm": 1.5890129050702244,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.7422567367553711,
983
+ "step": 1390
984
+ },
985
+ {
986
+ "epoch": 0.06352663581087213,
987
+ "grad_norm": 1.163231824259212,
988
+ "learning_rate": 2e-05,
989
+ "loss": 0.7530087947845459,
990
+ "step": 1400
991
+ },
992
+ {
993
+ "epoch": 0.0639803974952355,
994
+ "grad_norm": 1.9648397918825078,
995
+ "learning_rate": 2e-05,
996
+ "loss": 0.7386720180511475,
997
+ "step": 1410
998
+ },
999
+ {
1000
+ "epoch": 0.06443415917959887,
1001
+ "grad_norm": 1.1154033888302555,
1002
+ "learning_rate": 2e-05,
1003
+ "loss": 0.7690919399261474,
1004
+ "step": 1420
1005
+ },
1006
+ {
1007
+ "epoch": 0.06488792086396225,
1008
+ "grad_norm": 2.817335534651707,
1009
+ "learning_rate": 2e-05,
1010
+ "loss": 0.7828889846801758,
1011
+ "step": 1430
1012
+ },
1013
+ {
1014
+ "epoch": 0.06534168254832562,
1015
+ "grad_norm": 1.5652197356391113,
1016
+ "learning_rate": 2e-05,
1017
+ "loss": 0.7418618202209473,
1018
+ "step": 1440
1019
+ },
1020
+ {
1021
+ "epoch": 0.06579544423268899,
1022
+ "grad_norm": 1.4323144167320692,
1023
+ "learning_rate": 2e-05,
1024
+ "loss": 0.7751660823822022,
1025
+ "step": 1450
1026
+ },
1027
+ {
1028
+ "epoch": 0.06624920591705237,
1029
+ "grad_norm": 2.0558277244752623,
1030
+ "learning_rate": 2e-05,
1031
+ "loss": 0.7712644577026367,
1032
+ "step": 1460
1033
+ },
1034
+ {
1035
+ "epoch": 0.06670296760141574,
1036
+ "grad_norm": 19.677486710967138,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.7851467132568359,
1039
+ "step": 1470
1040
+ },
1041
+ {
1042
+ "epoch": 0.06715672928577911,
1043
+ "grad_norm": 25.384982473704017,
1044
+ "learning_rate": 2e-05,
1045
+ "loss": 0.7457910537719726,
1046
+ "step": 1480
1047
+ },
1048
+ {
1049
+ "epoch": 0.06761049097014248,
1050
+ "grad_norm": 1.1405705553998693,
1051
+ "learning_rate": 2e-05,
1052
+ "loss": 0.7223224639892578,
1053
+ "step": 1490
1054
+ },
1055
+ {
1056
+ "epoch": 0.06806425265450586,
1057
+ "grad_norm": 1.5986852504273663,
1058
+ "learning_rate": 2e-05,
1059
+ "loss": 0.7437433242797852,
1060
+ "step": 1500
1061
+ },
1062
+ {
1063
+ "epoch": 0.06851801433886923,
1064
+ "grad_norm": 1.473364803589886,
1065
+ "learning_rate": 2e-05,
1066
+ "loss": 0.7854445934295654,
1067
+ "step": 1510
1068
+ },
1069
+ {
1070
+ "epoch": 0.0689717760232326,
1071
+ "grad_norm": 10.816907441073242,
1072
+ "learning_rate": 2e-05,
1073
+ "loss": 0.7771403312683105,
1074
+ "step": 1520
1075
+ },
1076
+ {
1077
+ "epoch": 0.06942553770759598,
1078
+ "grad_norm": 2.3871999487587425,
1079
+ "learning_rate": 2e-05,
1080
+ "loss": 0.7488942623138428,
1081
+ "step": 1530
1082
+ },
1083
+ {
1084
+ "epoch": 0.06987929939195935,
1085
+ "grad_norm": 1.29078615018717,
1086
+ "learning_rate": 2e-05,
1087
+ "loss": 0.7702004432678222,
1088
+ "step": 1540
1089
+ },
1090
+ {
1091
+ "epoch": 0.07033306107632271,
1092
+ "grad_norm": 1.2427006882362452,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.7769896507263183,
1095
+ "step": 1550
1096
+ },
1097
+ {
1098
+ "epoch": 0.07078682276068608,
1099
+ "grad_norm": 1.8501499468782703,
1100
+ "learning_rate": 2e-05,
1101
+ "loss": 0.7520946025848388,
1102
+ "step": 1560
1103
+ },
1104
+ {
1105
+ "epoch": 0.07124058444504947,
1106
+ "grad_norm": 2.600210573565917,
1107
+ "learning_rate": 2e-05,
1108
+ "loss": 0.7633906841278076,
1109
+ "step": 1570
1110
+ },
1111
+ {
1112
+ "epoch": 0.07169434612941283,
1113
+ "grad_norm": 2.522289120850679,
1114
+ "learning_rate": 2e-05,
1115
+ "loss": 0.7302196502685547,
1116
+ "step": 1580
1117
+ },
1118
+ {
1119
+ "epoch": 0.0721481078137762,
1120
+ "grad_norm": 1.3515878124531273,
1121
+ "learning_rate": 2e-05,
1122
+ "loss": 0.7600698471069336,
1123
+ "step": 1590
1124
+ },
1125
+ {
1126
+ "epoch": 0.07260186949813957,
1127
+ "grad_norm": 1.140634184395665,
1128
+ "learning_rate": 2e-05,
1129
+ "loss": 0.7558602333068848,
1130
+ "step": 1600
1131
+ },
1132
+ {
1133
+ "epoch": 0.07305563118250295,
1134
+ "grad_norm": 1.3170969672293371,
1135
+ "learning_rate": 2e-05,
1136
+ "loss": 0.7792963027954102,
1137
+ "step": 1610
1138
+ },
1139
+ {
1140
+ "epoch": 0.07350939286686632,
1141
+ "grad_norm": 12.865316269450114,
1142
+ "learning_rate": 2e-05,
1143
+ "loss": 0.7376986980438233,
1144
+ "step": 1620
1145
+ },
1146
+ {
1147
+ "epoch": 0.07396315455122969,
1148
+ "grad_norm": 1.2524502323072844,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.7631961822509765,
1151
+ "step": 1630
1152
+ },
1153
+ {
1154
+ "epoch": 0.07441691623559307,
1155
+ "grad_norm": 1.1643893423939573,
1156
+ "learning_rate": 2e-05,
1157
+ "loss": 0.7555351257324219,
1158
+ "step": 1640
1159
+ },
1160
+ {
1161
+ "epoch": 0.07487067791995644,
1162
+ "grad_norm": 3.1366425147206507,
1163
+ "learning_rate": 2e-05,
1164
+ "loss": 0.7184365749359131,
1165
+ "step": 1650
1166
+ },
1167
+ {
1168
+ "epoch": 0.07532443960431981,
1169
+ "grad_norm": 1.6802684630791902,
1170
+ "learning_rate": 2e-05,
1171
+ "loss": 0.716264295578003,
1172
+ "step": 1660
1173
+ },
1174
+ {
1175
+ "epoch": 0.07577820128868318,
1176
+ "grad_norm": 1.4738450371157292,
1177
+ "learning_rate": 2e-05,
1178
+ "loss": 0.7608924865722656,
1179
+ "step": 1670
1180
+ },
1181
+ {
1182
+ "epoch": 0.07623196297304656,
1183
+ "grad_norm": 3.932860483703015,
1184
+ "learning_rate": 2e-05,
1185
+ "loss": 0.763948917388916,
1186
+ "step": 1680
1187
+ },
1188
+ {
1189
+ "epoch": 0.07668572465740993,
1190
+ "grad_norm": 1.6528466518745035,
1191
+ "learning_rate": 2e-05,
1192
+ "loss": 0.7618756771087647,
1193
+ "step": 1690
1194
+ },
1195
+ {
1196
+ "epoch": 0.0771394863417733,
1197
+ "grad_norm": 0.9245158103125077,
1198
+ "learning_rate": 2e-05,
1199
+ "loss": 0.7731586456298828,
1200
+ "step": 1700
1201
+ },
1202
+ {
1203
+ "epoch": 0.07759324802613668,
1204
+ "grad_norm": 1.3297115851518835,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.7360218048095704,
1207
+ "step": 1710
1208
+ },
1209
+ {
1210
+ "epoch": 0.07804700971050005,
1211
+ "grad_norm": 1.1975864742002036,
1212
+ "learning_rate": 2e-05,
1213
+ "loss": 0.7769705772399902,
1214
+ "step": 1720
1215
+ },
1216
+ {
1217
+ "epoch": 0.07850077139486342,
1218
+ "grad_norm": 1.2410115385284919,
1219
+ "learning_rate": 2e-05,
1220
+ "loss": 0.7371129035949707,
1221
+ "step": 1730
1222
+ },
1223
+ {
1224
+ "epoch": 0.07895453307922679,
1225
+ "grad_norm": 1.7317801478129846,
1226
+ "learning_rate": 2e-05,
1227
+ "loss": 0.7660929679870605,
1228
+ "step": 1740
1229
+ },
1230
+ {
1231
+ "epoch": 0.07940829476359017,
1232
+ "grad_norm": 0.9051550955007733,
1233
+ "learning_rate": 2e-05,
1234
+ "loss": 0.7519168376922607,
1235
+ "step": 1750
1236
+ },
1237
+ {
1238
+ "epoch": 0.07986205644795354,
1239
+ "grad_norm": 1.6723802945669413,
1240
+ "learning_rate": 2e-05,
1241
+ "loss": 0.7212412834167481,
1242
+ "step": 1760
1243
+ },
1244
+ {
1245
+ "epoch": 0.0803158181323169,
1246
+ "grad_norm": 0.9762152447317929,
1247
+ "learning_rate": 2e-05,
1248
+ "loss": 0.7753897666931152,
1249
+ "step": 1770
1250
+ },
1251
+ {
1252
+ "epoch": 0.08076957981668027,
1253
+ "grad_norm": 1.2473869795723174,
1254
+ "learning_rate": 2e-05,
1255
+ "loss": 0.7250448226928711,
1256
+ "step": 1780
1257
+ },
1258
+ {
1259
+ "epoch": 0.08122334150104366,
1260
+ "grad_norm": 3.1833001654859108,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.7206393241882324,
1263
+ "step": 1790
1264
+ },
1265
+ {
1266
+ "epoch": 0.08167710318540702,
1267
+ "grad_norm": 1.595724497195245,
1268
+ "learning_rate": 2e-05,
1269
+ "loss": 0.7571566581726075,
1270
+ "step": 1800
1271
+ },
1272
+ {
1273
+ "epoch": 0.08213086486977039,
1274
+ "grad_norm": 1.4572318958737844,
1275
+ "learning_rate": 2e-05,
1276
+ "loss": 0.7492036342620849,
1277
+ "step": 1810
1278
+ },
1279
+ {
1280
+ "epoch": 0.08258462655413377,
1281
+ "grad_norm": 2.696706475039584,
1282
+ "learning_rate": 2e-05,
1283
+ "loss": 0.757764196395874,
1284
+ "step": 1820
1285
+ },
1286
+ {
1287
+ "epoch": 0.08303838823849714,
1288
+ "grad_norm": 9.705539395813444,
1289
+ "learning_rate": 2e-05,
1290
+ "loss": 0.7491719245910644,
1291
+ "step": 1830
1292
+ },
1293
+ {
1294
+ "epoch": 0.08349214992286051,
1295
+ "grad_norm": 2.445157502577676,
1296
+ "learning_rate": 2e-05,
1297
+ "loss": 0.7391232967376709,
1298
+ "step": 1840
1299
+ },
1300
+ {
1301
+ "epoch": 0.08394591160722388,
1302
+ "grad_norm": 2.2218722227847745,
1303
+ "learning_rate": 2e-05,
1304
+ "loss": 0.7663413047790527,
1305
+ "step": 1850
1306
+ },
1307
+ {
1308
+ "epoch": 0.08439967329158726,
1309
+ "grad_norm": 6.437451692035588,
1310
+ "learning_rate": 2e-05,
1311
+ "loss": 0.7212537288665771,
1312
+ "step": 1860
1313
+ },
1314
+ {
1315
+ "epoch": 0.08485343497595063,
1316
+ "grad_norm": 14.459355718334681,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.7453588008880615,
1319
+ "step": 1870
1320
+ },
1321
+ {
1322
+ "epoch": 0.085307196660314,
1323
+ "grad_norm": 2.274387051434555,
1324
+ "learning_rate": 2e-05,
1325
+ "loss": 0.7336775779724121,
1326
+ "step": 1880
1327
+ },
1328
+ {
1329
+ "epoch": 0.08576095834467738,
1330
+ "grad_norm": 2.0275002345768827,
1331
+ "learning_rate": 2e-05,
1332
+ "loss": 0.8232375144958496,
1333
+ "step": 1890
1334
+ },
1335
+ {
1336
+ "epoch": 0.08621472002904075,
1337
+ "grad_norm": 3.4079005741269714,
1338
+ "learning_rate": 2e-05,
1339
+ "loss": 0.7306999206542969,
1340
+ "step": 1900
1341
+ },
1342
+ {
1343
+ "epoch": 0.08666848171340412,
1344
+ "grad_norm": 2.888480518708026,
1345
+ "learning_rate": 2e-05,
1346
+ "loss": 0.7341453552246093,
1347
+ "step": 1910
1348
+ },
1349
+ {
1350
+ "epoch": 0.08712224339776749,
1351
+ "grad_norm": 2.691347164562704,
1352
+ "learning_rate": 2e-05,
1353
+ "loss": 0.7440048217773437,
1354
+ "step": 1920
1355
+ },
1356
+ {
1357
+ "epoch": 0.08757600508213087,
1358
+ "grad_norm": 2.1392692969810208,
1359
+ "learning_rate": 2e-05,
1360
+ "loss": 0.7886202812194825,
1361
+ "step": 1930
1362
+ },
1363
+ {
1364
+ "epoch": 0.08802976676649424,
1365
+ "grad_norm": 3.1332404791177275,
1366
+ "learning_rate": 2e-05,
1367
+ "loss": 0.7390870571136474,
1368
+ "step": 1940
1369
+ },
1370
+ {
1371
+ "epoch": 0.08848352845085761,
1372
+ "grad_norm": 2.2360847842037592,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.7575728416442871,
1375
+ "step": 1950
1376
+ },
1377
+ {
1378
+ "epoch": 0.08893729013522098,
1379
+ "grad_norm": 4.969135746166459,
1380
+ "learning_rate": 2e-05,
1381
+ "loss": 0.734520435333252,
1382
+ "step": 1960
1383
+ },
1384
+ {
1385
+ "epoch": 0.08939105181958436,
1386
+ "grad_norm": 3.046212464054818,
1387
+ "learning_rate": 2e-05,
1388
+ "loss": 0.7530054092407227,
1389
+ "step": 1970
1390
+ },
1391
+ {
1392
+ "epoch": 0.08984481350394773,
1393
+ "grad_norm": 4.929848622369691,
1394
+ "learning_rate": 2e-05,
1395
+ "loss": 0.7513354301452637,
1396
+ "step": 1980
1397
+ },
1398
+ {
1399
+ "epoch": 0.0902985751883111,
1400
+ "grad_norm": 3.8207364141673184,
1401
+ "learning_rate": 2e-05,
1402
+ "loss": 0.7366059780120849,
1403
+ "step": 1990
1404
+ },
1405
+ {
1406
+ "epoch": 0.09075233687267448,
1407
+ "grad_norm": 3.445571619835256,
1408
+ "learning_rate": 2e-05,
1409
+ "loss": 0.7999159812927246,
1410
+ "step": 2000
1411
+ },
1412
+ {
1413
+ "epoch": 0.09120609855703785,
1414
+ "grad_norm": 2.210897089598567,
1415
+ "learning_rate": 2e-05,
1416
+ "loss": 0.7329919338226318,
1417
+ "step": 2010
1418
+ },
1419
+ {
1420
+ "epoch": 0.09165986024140121,
1421
+ "grad_norm": 1.3151585193661326,
1422
+ "learning_rate": 2e-05,
1423
+ "loss": 0.7405821800231933,
1424
+ "step": 2020
1425
+ },
1426
+ {
1427
+ "epoch": 0.09211362192576458,
1428
+ "grad_norm": 3.175239796553618,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.7426975250244141,
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 sha256:7c2396f2e8beedcfaa606e262219fba0791ddf36ad81eba0964a3e7c5f037e55
3
+ size 7313