Commit
•
9e8de2e
1
Parent(s):
4ea7f44
Training in progress, step 1500
Browse files- .gitattributes +1 -0
- .gitignore +1 -0
- config.json +282 -0
- create_model.py +59 -0
- decoder/config.json +74 -0
- decoder/pytorch_model.bin +3 -0
- merges.txt +0 -0
- preprocessor_config.json +9 -0
- pytorch_model.bin +3 -0
- run_librispeech.sh +36 -0
- run_speech_recognition_seq2seq.py +539 -0
- runs/Mar10_21-08-09_sanchit--v100/1646946516.4121053/events.out.tfevents.1646946516.sanchit--v100.1040434.1 +3 -0
- runs/Mar10_21-08-09_sanchit--v100/events.out.tfevents.1646946516.sanchit--v100.1040434.0 +3 -0
- runs/Mar10_21-09-23_sanchit--v100/1646946590.0115986/events.out.tfevents.1646946590.sanchit--v100.1043009.1 +3 -0
- runs/Mar10_21-09-23_sanchit--v100/events.out.tfevents.1646946590.sanchit--v100.1043009.0 +3 -0
- runs/Mar10_21-20-12_sanchit--v100/1646947238.8154027/events.out.tfevents.1646947238.sanchit--v100.1065811.1 +3 -0
- runs/Mar10_21-20-12_sanchit--v100/events.out.tfevents.1646947238.sanchit--v100.1065811.0 +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- training_args.bin +3 -0
- vocab.json +0 -0
- wandb/debug-internal.log +1 -0
- wandb/debug.log +1 -0
- wandb/latest-run +1 -0
- wandb/run-20220310_210836-1mziq47o/files/config.yaml +24 -0
- wandb/run-20220310_210836-1mziq47o/files/output.log +11 -0
- wandb/run-20220310_210836-1mziq47o/files/requirements.txt +184 -0
- wandb/run-20220310_210836-1mziq47o/files/wandb-metadata.json +59 -0
- wandb/run-20220310_210836-1mziq47o/files/wandb-summary.json +1 -0
- wandb/run-20220310_210836-1mziq47o/logs/debug-internal.log +57 -0
- wandb/run-20220310_210836-1mziq47o/logs/debug.log +29 -0
- wandb/run-20220310_210836-1mziq47o/run-1mziq47o.wandb +0 -0
- wandb/run-20220310_210950-19sg19ou/files/config.yaml +728 -0
- wandb/run-20220310_210950-19sg19ou/files/output.log +37 -0
- wandb/run-20220310_210950-19sg19ou/files/requirements.txt +184 -0
- wandb/run-20220310_210950-19sg19ou/files/wandb-metadata.json +59 -0
- wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json +1 -0
- wandb/run-20220310_210950-19sg19ou/logs/debug-internal.log +137 -0
- wandb/run-20220310_210950-19sg19ou/logs/debug.log +107 -0
- wandb/run-20220310_210950-19sg19ou/run-19sg19ou.wandb +0 -0
- wandb/run-20220310_212039-1nikmzrh/files/config.yaml +0 -0
- wandb/run-20220310_212039-1nikmzrh/files/output.log +0 -0
- wandb/run-20220310_212039-1nikmzrh/files/requirements.txt +184 -0
- wandb/run-20220310_212039-1nikmzrh/files/wandb-metadata.json +59 -0
- wandb/run-20220310_212039-1nikmzrh/files/wandb-summary.json +0 -0
- wandb/run-20220310_212039-1nikmzrh/logs/debug-internal.log +0 -0
- wandb/run-20220310_212039-1nikmzrh/logs/debug.log +27 -0
- wandb/run-20220310_212039-1nikmzrh/run-1nikmzrh.wandb +3 -0
.gitattributes
CHANGED
@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
25 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
25 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
28 |
+
wandb/run-20220310_212039-1nikmzrh/run-1nikmzrh.wandb filter=lfs diff=lfs merge=lfs -text
|
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
checkpoint-*/
|
config.json
ADDED
@@ -0,0 +1,282 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "./",
|
3 |
+
"architectures": [
|
4 |
+
"SpeechEncoderDecoderModel"
|
5 |
+
],
|
6 |
+
"decoder": {
|
7 |
+
"_name_or_path": "decoder",
|
8 |
+
"activation_dropout": 0.0,
|
9 |
+
"activation_function": "gelu",
|
10 |
+
"add_bias_logits": false,
|
11 |
+
"add_cross_attention": true,
|
12 |
+
"add_final_layer_norm": false,
|
13 |
+
"architectures": [
|
14 |
+
"BartForCausalLM"
|
15 |
+
],
|
16 |
+
"attention_dropout": 0.0,
|
17 |
+
"bad_words_ids": null,
|
18 |
+
"bos_token_id": 0,
|
19 |
+
"chunk_size_feed_forward": 0,
|
20 |
+
"classif_dropout": 0.0,
|
21 |
+
"classifier_dropout": 0.0,
|
22 |
+
"cross_attention_hidden_size": null,
|
23 |
+
"d_model": 1024,
|
24 |
+
"decoder_attention_heads": 16,
|
25 |
+
"decoder_ffn_dim": 4096,
|
26 |
+
"decoder_layerdrop": 0.0,
|
27 |
+
"decoder_layers": 12,
|
28 |
+
"decoder_start_token_id": 2,
|
29 |
+
"diversity_penalty": 0.0,
|
30 |
+
"do_sample": false,
|
31 |
+
"dropout": 0.0,
|
32 |
+
"early_stopping": true,
|
33 |
+
"encoder_attention_heads": 16,
|
34 |
+
"encoder_ffn_dim": 4096,
|
35 |
+
"encoder_layerdrop": 0.0,
|
36 |
+
"encoder_layers": 12,
|
37 |
+
"encoder_no_repeat_ngram_size": 0,
|
38 |
+
"eos_token_id": 2,
|
39 |
+
"finetuning_task": null,
|
40 |
+
"forced_bos_token_id": 0,
|
41 |
+
"forced_eos_token_id": 2,
|
42 |
+
"gradient_checkpointing": false,
|
43 |
+
"id2label": {
|
44 |
+
"0": "LABEL_0",
|
45 |
+
"1": "LABEL_1",
|
46 |
+
"2": "LABEL_2"
|
47 |
+
},
|
48 |
+
"init_std": 0.02,
|
49 |
+
"is_decoder": true,
|
50 |
+
"is_encoder_decoder": false,
|
51 |
+
"label2id": {
|
52 |
+
"LABEL_0": 0,
|
53 |
+
"LABEL_1": 1,
|
54 |
+
"LABEL_2": 2
|
55 |
+
},
|
56 |
+
"length_penalty": 1.0,
|
57 |
+
"max_length": 20,
|
58 |
+
"max_position_embeddings": 1024,
|
59 |
+
"min_length": 0,
|
60 |
+
"model_type": "bart",
|
61 |
+
"no_repeat_ngram_size": 3,
|
62 |
+
"normalize_before": false,
|
63 |
+
"num_beam_groups": 1,
|
64 |
+
"num_beams": 4,
|
65 |
+
"num_hidden_layers": 2,
|
66 |
+
"num_return_sequences": 1,
|
67 |
+
"output_attentions": false,
|
68 |
+
"output_hidden_states": false,
|
69 |
+
"output_scores": false,
|
70 |
+
"pad_token_id": 1,
|
71 |
+
"prefix": null,
|
72 |
+
"problem_type": null,
|
73 |
+
"pruned_heads": {},
|
74 |
+
"remove_invalid_values": false,
|
75 |
+
"repetition_penalty": 1.0,
|
76 |
+
"return_dict": true,
|
77 |
+
"return_dict_in_generate": false,
|
78 |
+
"scale_embedding": false,
|
79 |
+
"sep_token_id": null,
|
80 |
+
"task_specific_params": {
|
81 |
+
"summarization": {
|
82 |
+
"length_penalty": 1.0,
|
83 |
+
"max_length": 128,
|
84 |
+
"min_length": 12,
|
85 |
+
"num_beams": 4
|
86 |
+
},
|
87 |
+
"summarization_cnn": {
|
88 |
+
"length_penalty": 2.0,
|
89 |
+
"max_length": 142,
|
90 |
+
"min_length": 56,
|
91 |
+
"num_beams": 4
|
92 |
+
},
|
93 |
+
"summarization_xsum": {
|
94 |
+
"length_penalty": 1.0,
|
95 |
+
"max_length": 62,
|
96 |
+
"min_length": 11,
|
97 |
+
"num_beams": 6
|
98 |
+
}
|
99 |
+
},
|
100 |
+
"temperature": 1.0,
|
101 |
+
"tie_encoder_decoder": false,
|
102 |
+
"tie_word_embeddings": true,
|
103 |
+
"tokenizer_class": null,
|
104 |
+
"top_k": 50,
|
105 |
+
"top_p": 1.0,
|
106 |
+
"torch_dtype": "float32",
|
107 |
+
"torchscript": false,
|
108 |
+
"transformers_version": "4.17.0.dev0",
|
109 |
+
"use_bfloat16": false,
|
110 |
+
"use_cache": false,
|
111 |
+
"vocab_size": 50265
|
112 |
+
},
|
113 |
+
"decoder_start_token_id": 0,
|
114 |
+
"encoder": {
|
115 |
+
"_name_or_path": "facebook/wav2vec2-large-lv60",
|
116 |
+
"activation_dropout": 0.0,
|
117 |
+
"adapter_kernel_size": 3,
|
118 |
+
"adapter_stride": 2,
|
119 |
+
"add_adapter": true,
|
120 |
+
"add_cross_attention": false,
|
121 |
+
"apply_spec_augment": false,
|
122 |
+
"architectures": [
|
123 |
+
"Wav2Vec2ForPreTraining"
|
124 |
+
],
|
125 |
+
"attention_dropout": 0.0,
|
126 |
+
"bad_words_ids": null,
|
127 |
+
"bos_token_id": 1,
|
128 |
+
"chunk_size_feed_forward": 0,
|
129 |
+
"classifier_proj_size": 256,
|
130 |
+
"codevector_dim": 768,
|
131 |
+
"contrastive_logits_temperature": 0.1,
|
132 |
+
"conv_bias": true,
|
133 |
+
"conv_dim": [
|
134 |
+
512,
|
135 |
+
512,
|
136 |
+
512,
|
137 |
+
512,
|
138 |
+
512,
|
139 |
+
512,
|
140 |
+
512
|
141 |
+
],
|
142 |
+
"conv_kernel": [
|
143 |
+
10,
|
144 |
+
3,
|
145 |
+
3,
|
146 |
+
3,
|
147 |
+
3,
|
148 |
+
2,
|
149 |
+
2
|
150 |
+
],
|
151 |
+
"conv_stride": [
|
152 |
+
5,
|
153 |
+
2,
|
154 |
+
2,
|
155 |
+
2,
|
156 |
+
2,
|
157 |
+
2,
|
158 |
+
2
|
159 |
+
],
|
160 |
+
"cross_attention_hidden_size": null,
|
161 |
+
"ctc_loss_reduction": "sum",
|
162 |
+
"ctc_zero_infinity": false,
|
163 |
+
"decoder_start_token_id": null,
|
164 |
+
"diversity_loss_weight": 0.1,
|
165 |
+
"diversity_penalty": 0.0,
|
166 |
+
"do_sample": false,
|
167 |
+
"do_stable_layer_norm": true,
|
168 |
+
"early_stopping": false,
|
169 |
+
"encoder_no_repeat_ngram_size": 0,
|
170 |
+
"eos_token_id": 2,
|
171 |
+
"feat_extract_activation": "gelu",
|
172 |
+
"feat_extract_dropout": 0.0,
|
173 |
+
"feat_extract_norm": "layer",
|
174 |
+
"feat_proj_dropout": 0.0,
|
175 |
+
"feat_quantizer_dropout": 0.0,
|
176 |
+
"final_dropout": 0.0,
|
177 |
+
"finetuning_task": null,
|
178 |
+
"forced_bos_token_id": null,
|
179 |
+
"forced_eos_token_id": null,
|
180 |
+
"gradient_checkpointing": false,
|
181 |
+
"hidden_act": "gelu",
|
182 |
+
"hidden_dropout": 0.0,
|
183 |
+
"hidden_dropout_prob": 0.0,
|
184 |
+
"hidden_size": 1024,
|
185 |
+
"id2label": {
|
186 |
+
"0": "LABEL_0",
|
187 |
+
"1": "LABEL_1"
|
188 |
+
},
|
189 |
+
"initializer_range": 0.02,
|
190 |
+
"intermediate_size": 4096,
|
191 |
+
"is_decoder": false,
|
192 |
+
"is_encoder_decoder": false,
|
193 |
+
"label2id": {
|
194 |
+
"LABEL_0": 0,
|
195 |
+
"LABEL_1": 1
|
196 |
+
},
|
197 |
+
"layer_norm_eps": 1e-05,
|
198 |
+
"layerdrop": 0.0,
|
199 |
+
"length_penalty": 1.0,
|
200 |
+
"mask_feature_length": 10,
|
201 |
+
"mask_feature_min_masks": 0,
|
202 |
+
"mask_feature_prob": 0.0,
|
203 |
+
"mask_time_length": 10,
|
204 |
+
"mask_time_min_masks": 2,
|
205 |
+
"mask_time_prob": 0.0,
|
206 |
+
"max_length": 20,
|
207 |
+
"min_length": 0,
|
208 |
+
"model_type": "wav2vec2",
|
209 |
+
"no_repeat_ngram_size": 0,
|
210 |
+
"num_adapter_layers": 3,
|
211 |
+
"num_attention_heads": 16,
|
212 |
+
"num_beam_groups": 1,
|
213 |
+
"num_beams": 1,
|
214 |
+
"num_codevector_groups": 2,
|
215 |
+
"num_codevectors_per_group": 320,
|
216 |
+
"num_conv_pos_embedding_groups": 16,
|
217 |
+
"num_conv_pos_embeddings": 128,
|
218 |
+
"num_feat_extract_layers": 7,
|
219 |
+
"num_hidden_layers": 24,
|
220 |
+
"num_negatives": 100,
|
221 |
+
"num_return_sequences": 1,
|
222 |
+
"output_attentions": false,
|
223 |
+
"output_hidden_size": 1024,
|
224 |
+
"output_hidden_states": false,
|
225 |
+
"output_scores": false,
|
226 |
+
"pad_token_id": 0,
|
227 |
+
"prefix": null,
|
228 |
+
"problem_type": null,
|
229 |
+
"proj_codevector_dim": 768,
|
230 |
+
"pruned_heads": {},
|
231 |
+
"remove_invalid_values": false,
|
232 |
+
"repetition_penalty": 1.0,
|
233 |
+
"return_dict": true,
|
234 |
+
"return_dict_in_generate": false,
|
235 |
+
"sep_token_id": null,
|
236 |
+
"task_specific_params": null,
|
237 |
+
"tdnn_dilation": [
|
238 |
+
1,
|
239 |
+
2,
|
240 |
+
3,
|
241 |
+
1,
|
242 |
+
1
|
243 |
+
],
|
244 |
+
"tdnn_dim": [
|
245 |
+
512,
|
246 |
+
512,
|
247 |
+
512,
|
248 |
+
512,
|
249 |
+
1500
|
250 |
+
],
|
251 |
+
"tdnn_kernel": [
|
252 |
+
5,
|
253 |
+
3,
|
254 |
+
3,
|
255 |
+
1,
|
256 |
+
1
|
257 |
+
],
|
258 |
+
"temperature": 1.0,
|
259 |
+
"tie_encoder_decoder": false,
|
260 |
+
"tie_word_embeddings": true,
|
261 |
+
"tokenizer_class": null,
|
262 |
+
"top_k": 50,
|
263 |
+
"top_p": 1.0,
|
264 |
+
"torch_dtype": null,
|
265 |
+
"torchscript": false,
|
266 |
+
"transformers_version": "4.17.0.dev0",
|
267 |
+
"use_bfloat16": false,
|
268 |
+
"use_weighted_layer_sum": false,
|
269 |
+
"vocab_size": 32,
|
270 |
+
"xvector_output_dim": 512
|
271 |
+
},
|
272 |
+
"eos_token_id": 2,
|
273 |
+
"is_encoder_decoder": true,
|
274 |
+
"max_length": 50,
|
275 |
+
"model_type": "speech-encoder-decoder",
|
276 |
+
"pad_token_id": 1,
|
277 |
+
"processor_class": "Wav2Vec2Processor",
|
278 |
+
"tie_word_embeddings": false,
|
279 |
+
"torch_dtype": "float32",
|
280 |
+
"transformers_version": null,
|
281 |
+
"use_cache": false
|
282 |
+
}
|
create_model.py
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import SpeechEncoderDecoderModel, AutoFeatureExtractor, AutoTokenizer, Wav2Vec2Processor, BartConfig, BartForCausalLM
|
2 |
+
import torch
|
3 |
+
|
4 |
+
# checkpoints to leverage
|
5 |
+
encoder_id = "facebook/wav2vec2-large-lv60"
|
6 |
+
decoder_id = "facebook/bart-large"
|
7 |
+
|
8 |
+
feature_extractor = AutoFeatureExtractor.from_pretrained(encoder_id)
|
9 |
+
feature_extractor.save_pretrained("./")
|
10 |
+
tokenizer = AutoTokenizer.from_pretrained(decoder_id)
|
11 |
+
tokenizer.save_pretrained("./")
|
12 |
+
|
13 |
+
decoder_config = BartConfig.from_pretrained(decoder_id, is_decoder=True)
|
14 |
+
decoder_config.num_hidden_layers = 2
|
15 |
+
decoder = BartForCausalLM(decoder_config)
|
16 |
+
decoder.save_pretrained("decoder") # save the decoder
|
17 |
+
|
18 |
+
model = SpeechEncoderDecoderModel.from_encoder_decoder_pretrained(encoder_id, "decoder", encoder_add_adapter=True)
|
19 |
+
model.config.encoder.feat_proj_dropout = 0.0
|
20 |
+
model.config.encoder.final_dropout = 0.0
|
21 |
+
model.config.encoder.mask_time_prob = 0.1
|
22 |
+
model.config.decoder_start_token_id = model.decoder.config.bos_token_id
|
23 |
+
model.config.pad_token_id = model.decoder.config.pad_token_id
|
24 |
+
model.config.eos_token_id = model.decoder.config.eos_token_id
|
25 |
+
model.config.max_length = 50
|
26 |
+
model.config.num_beams = 1
|
27 |
+
model.config.encoder.layerdrop = 0.0
|
28 |
+
model.config.use_cache = False
|
29 |
+
model.config.decoder.use_cache = False
|
30 |
+
model.config.processor_class = "Wav2Vec2Processor"
|
31 |
+
|
32 |
+
# set all encoder regularisation to zero
|
33 |
+
model.config.encoder.feat_proj_dropout = 0.0
|
34 |
+
model.config.encoder.final_dropout = 0.0
|
35 |
+
model.config.encoder.activation_dropout = 0.0
|
36 |
+
model.config.encoder.apply_spec_augment = False
|
37 |
+
model.config.encoder.attention_dropout = 0.0
|
38 |
+
model.config.encoder.feat_extract_dropout = 0.0
|
39 |
+
model.config.encoder.feat_proj_dropout = 0.0
|
40 |
+
model.config.encoder.hidden_dropout = 0.0
|
41 |
+
model.config.encoder.hidden_dropout_prob = 0.0
|
42 |
+
model.config.encoder.layerdrop = 0.0
|
43 |
+
model.config.encoder.mask_feature_prob = 0.0
|
44 |
+
model.config.encoder.mask_time_prob = 0.0
|
45 |
+
|
46 |
+
|
47 |
+
# set all decoder regularisation to zero
|
48 |
+
model.config.decoder.activation_dropout = 0.0
|
49 |
+
model.config.decoder.attention_dropout = 0.0
|
50 |
+
model.config.decoder.classif_dropout = 0.0
|
51 |
+
model.config.decoder.classifier_dropout = 0.0
|
52 |
+
model.config.decoder.dropout = 0.0
|
53 |
+
|
54 |
+
# check if generation works
|
55 |
+
out = model.generate(torch.ones((1, 2000)))
|
56 |
+
|
57 |
+
model.save_pretrained("./")
|
58 |
+
|
59 |
+
|
decoder/config.json
ADDED
@@ -0,0 +1,74 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"activation_dropout": 0.1,
|
3 |
+
"activation_function": "gelu",
|
4 |
+
"add_bias_logits": false,
|
5 |
+
"add_final_layer_norm": false,
|
6 |
+
"architectures": [
|
7 |
+
"BartForCausalLM"
|
8 |
+
],
|
9 |
+
"attention_dropout": 0.1,
|
10 |
+
"bos_token_id": 0,
|
11 |
+
"classif_dropout": 0.1,
|
12 |
+
"classifier_dropout": 0.0,
|
13 |
+
"d_model": 1024,
|
14 |
+
"decoder_attention_heads": 16,
|
15 |
+
"decoder_ffn_dim": 4096,
|
16 |
+
"decoder_layerdrop": 0.0,
|
17 |
+
"decoder_layers": 12,
|
18 |
+
"decoder_start_token_id": 2,
|
19 |
+
"dropout": 0.1,
|
20 |
+
"early_stopping": true,
|
21 |
+
"encoder_attention_heads": 16,
|
22 |
+
"encoder_ffn_dim": 4096,
|
23 |
+
"encoder_layerdrop": 0.0,
|
24 |
+
"encoder_layers": 12,
|
25 |
+
"eos_token_id": 2,
|
26 |
+
"forced_bos_token_id": 0,
|
27 |
+
"forced_eos_token_id": 2,
|
28 |
+
"gradient_checkpointing": false,
|
29 |
+
"id2label": {
|
30 |
+
"0": "LABEL_0",
|
31 |
+
"1": "LABEL_1",
|
32 |
+
"2": "LABEL_2"
|
33 |
+
},
|
34 |
+
"init_std": 0.02,
|
35 |
+
"is_decoder": true,
|
36 |
+
"is_encoder_decoder": false,
|
37 |
+
"label2id": {
|
38 |
+
"LABEL_0": 0,
|
39 |
+
"LABEL_1": 1,
|
40 |
+
"LABEL_2": 2
|
41 |
+
},
|
42 |
+
"max_position_embeddings": 1024,
|
43 |
+
"model_type": "bart",
|
44 |
+
"no_repeat_ngram_size": 3,
|
45 |
+
"normalize_before": false,
|
46 |
+
"num_beams": 4,
|
47 |
+
"num_hidden_layers": 2,
|
48 |
+
"pad_token_id": 1,
|
49 |
+
"scale_embedding": false,
|
50 |
+
"task_specific_params": {
|
51 |
+
"summarization": {
|
52 |
+
"length_penalty": 1.0,
|
53 |
+
"max_length": 128,
|
54 |
+
"min_length": 12,
|
55 |
+
"num_beams": 4
|
56 |
+
},
|
57 |
+
"summarization_cnn": {
|
58 |
+
"length_penalty": 2.0,
|
59 |
+
"max_length": 142,
|
60 |
+
"min_length": 56,
|
61 |
+
"num_beams": 4
|
62 |
+
},
|
63 |
+
"summarization_xsum": {
|
64 |
+
"length_penalty": 1.0,
|
65 |
+
"max_length": 62,
|
66 |
+
"min_length": 11,
|
67 |
+
"num_beams": 6
|
68 |
+
}
|
69 |
+
},
|
70 |
+
"torch_dtype": "float32",
|
71 |
+
"transformers_version": "4.17.0.dev0",
|
72 |
+
"use_cache": true,
|
73 |
+
"vocab_size": 50265
|
74 |
+
}
|
decoder/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:57f62bd9371551a17276870b6a8b042e7bba02a89ec86aadac2ee1e1409ab0db
|
3 |
+
size 1016449417
|
merges.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
preprocessor_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
4 |
+
"feature_size": 1,
|
5 |
+
"padding_side": "right",
|
6 |
+
"padding_value": 0.0,
|
7 |
+
"return_attention_mask": true,
|
8 |
+
"sampling_rate": 16000
|
9 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3892e4489e781e4a6f5e07bc6ce6a57d3def4573edfe203df67798d5a448acf9
|
3 |
+
size 2353901299
|
run_librispeech.sh
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/usr/bin/env bash
|
2 |
+
CUDA_VISIBLE_DEVICES=1 python run_speech_recognition_seq2seq.py \
|
3 |
+
--dataset_name="librispeech_asr" \
|
4 |
+
--model_name_or_path="./" \
|
5 |
+
--tokenizer_name="./" \
|
6 |
+
--dataset_config_name="clean" \
|
7 |
+
--train_split_name="train.100" \
|
8 |
+
--eval_split_name="validation" \
|
9 |
+
--output_dir="./" \
|
10 |
+
--preprocessing_num_workers="1" \
|
11 |
+
--length_column_name="input_length" \
|
12 |
+
--overwrite_output_dir \
|
13 |
+
--num_train_epochs="20" \
|
14 |
+
--per_device_train_batch_size="8" \
|
15 |
+
--per_device_eval_batch_size="8" \
|
16 |
+
--gradient_accumulation_steps="4" \
|
17 |
+
--generation_max_length="40" \
|
18 |
+
--generation_num_beams="1" \
|
19 |
+
--learning_rate="3e-5" \
|
20 |
+
--warmup_steps="1000" \
|
21 |
+
--evaluation_strategy="steps" \
|
22 |
+
--text_column_name="text" \
|
23 |
+
--save_steps="1500" \
|
24 |
+
--eval_steps="1500" \
|
25 |
+
--logging_steps="1" \
|
26 |
+
--save_total_limit="1" \
|
27 |
+
--freeze_feature_encoder \
|
28 |
+
--gradient_checkpointing \
|
29 |
+
--fp16 \
|
30 |
+
--group_by_length \
|
31 |
+
--predict_with_generate \
|
32 |
+
--do_lower_case \
|
33 |
+
--do_eval --do_train \
|
34 |
+
--push_to_hub \
|
35 |
+
--use_auth_token
|
36 |
+
|
run_speech_recognition_seq2seq.py
ADDED
@@ -0,0 +1,539 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#!/usr/bin/env python
|
2 |
+
# coding=utf-8
|
3 |
+
# Copyright 2021 The HuggingFace Team. All rights reserved.
|
4 |
+
#
|
5 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
6 |
+
# you may not use this file except in compliance with the License.
|
7 |
+
# You may obtain a copy of the License at
|
8 |
+
#
|
9 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
10 |
+
#
|
11 |
+
# Unless required by applicable law or agreed to in writing, software
|
12 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
13 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
14 |
+
# See the License for the specific language governing permissions and
|
15 |
+
# limitations under the License.
|
16 |
+
"""
|
17 |
+
Fine-tuning the library models for sequence to sequence speech recognition.
|
18 |
+
"""
|
19 |
+
# You can also adapt this script on your own sequence to sequence speech
|
20 |
+
# recognition task. Pointers for this are left as comments.
|
21 |
+
|
22 |
+
import logging
|
23 |
+
import os
|
24 |
+
import sys
|
25 |
+
from dataclasses import dataclass, field
|
26 |
+
from typing import Any, Dict, List, Optional, Union
|
27 |
+
|
28 |
+
import datasets
|
29 |
+
import torch
|
30 |
+
from datasets import DatasetDict, load_dataset, load_metric
|
31 |
+
|
32 |
+
import bitsandbytes as bnb
|
33 |
+
import transformers
|
34 |
+
from transformers import (
|
35 |
+
AutoConfig,
|
36 |
+
AutoFeatureExtractor,
|
37 |
+
AutoModelForSpeechSeq2Seq,
|
38 |
+
AutoProcessor,
|
39 |
+
AutoTokenizer,
|
40 |
+
HfArgumentParser,
|
41 |
+
Seq2SeqTrainer,
|
42 |
+
Seq2SeqTrainingArguments,
|
43 |
+
set_seed,
|
44 |
+
)
|
45 |
+
from transformers.trainer_pt_utils import get_parameter_names
|
46 |
+
from transformers.trainer_utils import get_last_checkpoint, is_main_process
|
47 |
+
from transformers.utils import check_min_version
|
48 |
+
from transformers.utils.versions import require_version
|
49 |
+
from transformers.optimization import Adafactor
|
50 |
+
|
51 |
+
|
52 |
+
# Will error if the minimal version of Transformers is not installed. Remove at your own risks.
|
53 |
+
check_min_version("4.17.0.dev0")
|
54 |
+
|
55 |
+
require_version("datasets>=1.8.0", "To fix: pip install -r examples/pytorch/summarization/requirements.txt")
|
56 |
+
|
57 |
+
logger = logging.getLogger(__name__)
|
58 |
+
|
59 |
+
|
60 |
+
@dataclass
|
61 |
+
class ModelArguments:
|
62 |
+
"""
|
63 |
+
Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
|
64 |
+
"""
|
65 |
+
|
66 |
+
model_name_or_path: str = field(
|
67 |
+
metadata={"help": "Path to pretrained model or model identifier from huggingface.co/models"}
|
68 |
+
)
|
69 |
+
config_name: Optional[str] = field(
|
70 |
+
default=None, metadata={"help": "Pretrained config name or path if not the same as model_name"}
|
71 |
+
)
|
72 |
+
tokenizer_name: Optional[str] = field(
|
73 |
+
default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
|
74 |
+
)
|
75 |
+
feature_extractor_name: Optional[str] = field(
|
76 |
+
default=None, metadata={"help": "feature extractor name or path if not the same as model_name"}
|
77 |
+
)
|
78 |
+
cache_dir: Optional[str] = field(
|
79 |
+
default=None,
|
80 |
+
metadata={"help": "Where to store the pretrained models downloaded from huggingface.co"},
|
81 |
+
)
|
82 |
+
use_fast_tokenizer: bool = field(
|
83 |
+
default=True,
|
84 |
+
metadata={"help": "Whether to use one of the fast tokenizer (backed by the tokenizers library) or not."},
|
85 |
+
)
|
86 |
+
model_revision: str = field(
|
87 |
+
default="main",
|
88 |
+
metadata={"help": "The specific model version to use (can be a branch name, tag name or commit id)."},
|
89 |
+
)
|
90 |
+
use_auth_token: bool = field(
|
91 |
+
default=False,
|
92 |
+
metadata={
|
93 |
+
"help": "Will use the token generated when running `transformers-cli login` (necessary to use this script "
|
94 |
+
"with private models)."
|
95 |
+
},
|
96 |
+
)
|
97 |
+
freeze_feature_encoder: bool = field(
|
98 |
+
default=True, metadata={"help": "Whether to freeze the feature encoder layers of the model."}
|
99 |
+
)
|
100 |
+
|
101 |
+
|
102 |
+
@dataclass
|
103 |
+
class DataTrainingArguments:
|
104 |
+
"""
|
105 |
+
Arguments pertaining to what data we are going to input our model for training and eval.
|
106 |
+
"""
|
107 |
+
|
108 |
+
dataset_name: str = field(
|
109 |
+
default=None, metadata={"help": "The name of the dataset to use (via the datasets library)."}
|
110 |
+
)
|
111 |
+
dataset_config_name: Optional[str] = field(
|
112 |
+
default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
|
113 |
+
)
|
114 |
+
text_column: Optional[str] = field(
|
115 |
+
default=None,
|
116 |
+
metadata={"help": "The name of the column in the datasets containing the full texts (for summarization)."},
|
117 |
+
)
|
118 |
+
overwrite_cache: bool = field(
|
119 |
+
default=False, metadata={"help": "Overwrite the cached training and evaluation sets"}
|
120 |
+
)
|
121 |
+
preprocessing_num_workers: Optional[int] = field(
|
122 |
+
default=None,
|
123 |
+
metadata={"help": "The number of processes to use for the preprocessing."},
|
124 |
+
)
|
125 |
+
max_train_samples: Optional[int] = field(
|
126 |
+
default=None,
|
127 |
+
metadata={
|
128 |
+
"help": "For debugging purposes or quicker training, truncate the number of training examples to this "
|
129 |
+
"value if set."
|
130 |
+
},
|
131 |
+
)
|
132 |
+
max_eval_samples: Optional[int] = field(
|
133 |
+
default=None,
|
134 |
+
metadata={
|
135 |
+
"help": "For debugging purposes or quicker training, truncate the number of evaluation examples to this "
|
136 |
+
"value if set."
|
137 |
+
},
|
138 |
+
)
|
139 |
+
audio_column_name: str = field(
|
140 |
+
default="audio",
|
141 |
+
metadata={"help": "The name of the dataset column containing the audio data. Defaults to 'audio'"},
|
142 |
+
)
|
143 |
+
text_column_name: str = field(
|
144 |
+
default="text",
|
145 |
+
metadata={"help": "The name of the dataset column containing the text data. Defaults to 'text'"},
|
146 |
+
)
|
147 |
+
max_duration_in_seconds: float = field(
|
148 |
+
default=20.0,
|
149 |
+
metadata={
|
150 |
+
"help": "Truncate audio files that are longer than `max_duration_in_seconds` seconds to 'max_duration_in_seconds`"
|
151 |
+
},
|
152 |
+
)
|
153 |
+
min_duration_in_seconds: float = field(
|
154 |
+
default=0.0, metadata={"help": "Filter audio files that are shorter than `min_duration_in_seconds` seconds"}
|
155 |
+
)
|
156 |
+
preprocessing_only: bool = field(
|
157 |
+
default=False,
|
158 |
+
metadata={
|
159 |
+
"help": "Whether to only do data preprocessing and skip training. "
|
160 |
+
"This is especially useful when data preprocessing errors out in distributed training due to timeout. "
|
161 |
+
"In this case, one should run the preprocessing in a non-distributed setup with `preprocessing_only=True` "
|
162 |
+
"so that the cached datasets can consequently be loaded in distributed training"
|
163 |
+
},
|
164 |
+
)
|
165 |
+
train_split_name: str = field(
|
166 |
+
default="train",
|
167 |
+
metadata={
|
168 |
+
"help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
|
169 |
+
},
|
170 |
+
)
|
171 |
+
eval_split_name: str = field(
|
172 |
+
default="test",
|
173 |
+
metadata={
|
174 |
+
"help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
|
175 |
+
},
|
176 |
+
)
|
177 |
+
do_lower_case: bool = field(
|
178 |
+
default=True,
|
179 |
+
metadata={"help": "Whether the target text should be lower cased."},
|
180 |
+
)
|
181 |
+
|
182 |
+
|
183 |
+
@dataclass
|
184 |
+
class DataCollatorSpeechSeq2SeqWithPadding:
|
185 |
+
"""
|
186 |
+
Data collator that will dynamically pad the inputs received.
|
187 |
+
Args:
|
188 |
+
processor ([`Wav2Vec2Processor`])
|
189 |
+
The processor used for proccessing the data.
|
190 |
+
decoder_start_token_id (`int`)
|
191 |
+
The begin-of-sentence of the decoder.
|
192 |
+
"""
|
193 |
+
|
194 |
+
processor: Any
|
195 |
+
decoder_start_token_id: int
|
196 |
+
|
197 |
+
def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
|
198 |
+
# split inputs and labels since they have to be of different lenghts and need
|
199 |
+
# different padding methods
|
200 |
+
input_features = [{"input_values": feature["input_values"]} for feature in features]
|
201 |
+
label_features = [{"input_ids": feature["labels"]} for feature in features]
|
202 |
+
|
203 |
+
batch = self.processor.feature_extractor.pad(input_features, return_tensors="pt")
|
204 |
+
|
205 |
+
labels_batch = self.processor.tokenizer.pad(label_features, return_tensors="pt")
|
206 |
+
|
207 |
+
# replace padding with -100 to ignore loss correctly
|
208 |
+
labels = labels_batch["input_ids"].masked_fill(labels_batch.attention_mask.ne(1), -100)
|
209 |
+
|
210 |
+
# if bos token is appended in previous tokenization step,
|
211 |
+
# cut bos token here as it's append later anyways
|
212 |
+
if (labels[:, 0] == self.decoder_start_token_id).all().cpu().item():
|
213 |
+
labels = labels[:, 1:]
|
214 |
+
|
215 |
+
batch["labels"] = labels
|
216 |
+
|
217 |
+
return batch
|
218 |
+
|
219 |
+
|
220 |
+
def main():
|
221 |
+
# 1. Parse input arguments
|
222 |
+
# See all possible arguments in src/transformers/training_args.py
|
223 |
+
# or by passing the --help flag to this script.
|
224 |
+
# We now keep distinct sets of args, for a cleaner separation of concerns.
|
225 |
+
parser = HfArgumentParser((ModelArguments, DataTrainingArguments, Seq2SeqTrainingArguments))
|
226 |
+
|
227 |
+
if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
|
228 |
+
# If we pass only one argument to the script and it's the path to a json file,
|
229 |
+
# let's parse it to get our arguments.
|
230 |
+
model_args, data_args, training_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))
|
231 |
+
else:
|
232 |
+
model_args, data_args, training_args = parser.parse_args_into_dataclasses()
|
233 |
+
|
234 |
+
# 2. Setup logging
|
235 |
+
logging.basicConfig(
|
236 |
+
format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
|
237 |
+
datefmt="%m/%d/%Y %H:%M:%S",
|
238 |
+
handlers=[logging.StreamHandler(sys.stdout)],
|
239 |
+
)
|
240 |
+
log_level = training_args.get_process_log_level()
|
241 |
+
logger.setLevel(log_level)
|
242 |
+
datasets.utils.logging.set_verbosity(log_level)
|
243 |
+
transformers.utils.logging.set_verbosity(log_level)
|
244 |
+
transformers.utils.logging.enable_default_handler()
|
245 |
+
transformers.utils.logging.enable_explicit_format()
|
246 |
+
|
247 |
+
logger.setLevel(logging.INFO if is_main_process(training_args.local_rank) else logging.WARN)
|
248 |
+
|
249 |
+
# Log on each process the small summary:
|
250 |
+
logger.warning(
|
251 |
+
f"Process rank: {training_args.local_rank}, device: {training_args.device}, n_gpu: {training_args.n_gpu}"
|
252 |
+
f"distributed training: {bool(training_args.local_rank != -1)}, 16-bits training: {training_args.fp16}"
|
253 |
+
)
|
254 |
+
logger.info(f"Training/evaluation parameters {training_args}")
|
255 |
+
|
256 |
+
# Set the verbosity to info of the Transformers logger (on main process only):
|
257 |
+
if is_main_process(training_args.local_rank):
|
258 |
+
transformers.utils.logging.set_verbosity_info()
|
259 |
+
logger.info("Training/evaluation parameters %s", training_args)
|
260 |
+
|
261 |
+
# 3. Detecting last checkpoint and eventualy continue from last checkpoint
|
262 |
+
last_checkpoint = None
|
263 |
+
if os.path.isdir(training_args.output_dir) and training_args.do_train and not training_args.overwrite_output_dir:
|
264 |
+
last_checkpoint = get_last_checkpoint(training_args.output_dir)
|
265 |
+
if last_checkpoint is None and len(os.listdir(training_args.output_dir)) > 0:
|
266 |
+
raise ValueError(
|
267 |
+
f"Output directory ({training_args.output_dir}) already exists and is not empty. "
|
268 |
+
"Use --overwrite_output_dir to overcome."
|
269 |
+
)
|
270 |
+
elif last_checkpoint is not None and training_args.resume_from_checkpoint is None:
|
271 |
+
logger.info(
|
272 |
+
f"Checkpoint detected, resuming training at {last_checkpoint}. To avoid this behavior, change "
|
273 |
+
"the `--output_dir` or add `--overwrite_output_dir` to train from scratch."
|
274 |
+
)
|
275 |
+
|
276 |
+
# Set seed before initializing model.
|
277 |
+
set_seed(training_args.seed)
|
278 |
+
|
279 |
+
# 4. Load dataset
|
280 |
+
raw_datasets = DatasetDict()
|
281 |
+
|
282 |
+
if training_args.do_train:
|
283 |
+
raw_datasets["train"] = load_dataset(
|
284 |
+
data_args.dataset_name, data_args.dataset_config_name, split=data_args.train_split_name
|
285 |
+
)
|
286 |
+
|
287 |
+
if training_args.do_eval:
|
288 |
+
raw_datasets["eval"] = load_dataset(
|
289 |
+
data_args.dataset_name, data_args.dataset_config_name, split=data_args.eval_split_name
|
290 |
+
)
|
291 |
+
|
292 |
+
if data_args.audio_column_name not in next(iter(raw_datasets.values())).column_names:
|
293 |
+
raise ValueError(
|
294 |
+
f"--audio_column_name '{data_args.audio_column_name}' not found in dataset '{data_args.dataset_name}'. "
|
295 |
+
"Make sure to set `--audio_column_name` to the correct audio column - one of "
|
296 |
+
f"{', '.join(next(iter(raw_datasets.values())).column_names)}."
|
297 |
+
)
|
298 |
+
|
299 |
+
if data_args.text_column_name not in next(iter(raw_datasets.values())).column_names:
|
300 |
+
raise ValueError(
|
301 |
+
f"--text_column_name {data_args.text_column_name} not found in dataset '{data_args.dataset_name}'. "
|
302 |
+
"Make sure to set `--text_column_name` to the correct text column - one of "
|
303 |
+
f"{', '.join(next(iter(raw_datasets.values())).column_names)}."
|
304 |
+
)
|
305 |
+
|
306 |
+
# 5. Load pretrained model, tokenizer, and feature extractor
|
307 |
+
#
|
308 |
+
# Distributed training:
|
309 |
+
# The .from_pretrained methods guarantee that only one local process can concurrently
|
310 |
+
config = AutoConfig.from_pretrained(
|
311 |
+
model_args.config_name if model_args.config_name else model_args.model_name_or_path,
|
312 |
+
cache_dir=model_args.cache_dir,
|
313 |
+
revision=model_args.model_revision,
|
314 |
+
use_auth_token=True if model_args.use_auth_token else None,
|
315 |
+
)
|
316 |
+
|
317 |
+
feature_extractor = AutoFeatureExtractor.from_pretrained(
|
318 |
+
model_args.feature_extractor_name if model_args.feature_extractor_name else model_args.model_name_or_path,
|
319 |
+
cache_dir=model_args.cache_dir,
|
320 |
+
revision=model_args.model_revision,
|
321 |
+
use_auth_token=True if model_args.use_auth_token else None,
|
322 |
+
)
|
323 |
+
tokenizer = AutoTokenizer.from_pretrained(
|
324 |
+
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
|
325 |
+
cache_dir=model_args.cache_dir,
|
326 |
+
use_fast=model_args.use_fast_tokenizer,
|
327 |
+
revision=model_args.model_revision,
|
328 |
+
use_auth_token=True if model_args.use_auth_token else None,
|
329 |
+
)
|
330 |
+
model = AutoModelForSpeechSeq2Seq.from_pretrained(
|
331 |
+
model_args.model_name_or_path,
|
332 |
+
config=config,
|
333 |
+
cache_dir=model_args.cache_dir,
|
334 |
+
revision=model_args.model_revision,
|
335 |
+
use_auth_token=True if model_args.use_auth_token else None,
|
336 |
+
)
|
337 |
+
|
338 |
+
if model.config.decoder_start_token_id is None:
|
339 |
+
raise ValueError("Make sure that `config.decoder_start_token_id` is correctly defined")
|
340 |
+
|
341 |
+
if model_args.freeze_feature_encoder:
|
342 |
+
model.freeze_feature_encoder()
|
343 |
+
|
344 |
+
# 6. Resample speech dataset if necassary
|
345 |
+
dataset_sampling_rate = next(iter(raw_datasets.values())).features[data_args.audio_column_name].sampling_rate
|
346 |
+
if dataset_sampling_rate != feature_extractor.sampling_rate:
|
347 |
+
raw_datasets = raw_datasets.cast_column(
|
348 |
+
data_args.audio_column_name, datasets.features.Audio(sampling_rate=feature_extractor.sampling_rate)
|
349 |
+
)
|
350 |
+
|
351 |
+
# 7. Preprocessing the datasets.
|
352 |
+
# We need to read the audio files as arrays and tokenize the targets.
|
353 |
+
max_input_length = data_args.max_duration_in_seconds * feature_extractor.sampling_rate
|
354 |
+
min_input_length = data_args.min_duration_in_seconds * feature_extractor.sampling_rate
|
355 |
+
audio_column_name = data_args.audio_column_name
|
356 |
+
num_workers = data_args.preprocessing_num_workers
|
357 |
+
text_column_name = data_args.text_column_name
|
358 |
+
model_input_name = feature_extractor.model_input_names[0]
|
359 |
+
do_lower_case = data_args.do_lower_case
|
360 |
+
|
361 |
+
if data_args.max_train_samples is not None:
|
362 |
+
raw_datasets["train"] = raw_datasets["train"].select(range(data_args.max_train_samples))
|
363 |
+
|
364 |
+
if data_args.max_eval_samples is not None:
|
365 |
+
raw_datasets["eval"] = raw_datasets["eval"].select(range(data_args.max_eval_samples))
|
366 |
+
|
367 |
+
def prepare_dataset(batch):
|
368 |
+
# process audio
|
369 |
+
sample = batch[audio_column_name]
|
370 |
+
inputs = feature_extractor(sample["array"], sampling_rate=sample["sampling_rate"])
|
371 |
+
# process audio length
|
372 |
+
batch[model_input_name] = inputs.input_values[0]
|
373 |
+
batch["input_length"] = len(batch["input_values"])
|
374 |
+
|
375 |
+
# process targets
|
376 |
+
input_str = batch[text_column_name].lower() if do_lower_case else batch[text_column_name]
|
377 |
+
batch["labels"] = tokenizer(input_str).input_ids
|
378 |
+
return batch
|
379 |
+
|
380 |
+
with training_args.main_process_first(desc="dataset map pre-processing"):
|
381 |
+
vectorized_datasets = raw_datasets.map(
|
382 |
+
prepare_dataset,
|
383 |
+
remove_columns=next(iter(raw_datasets.values())).column_names,
|
384 |
+
num_proc=data_args.preprocessing_num_workers,
|
385 |
+
desc="preprocess train dataset",
|
386 |
+
)
|
387 |
+
|
388 |
+
# filter data that is shorter than min_input_length or longer than
|
389 |
+
# max_input_length
|
390 |
+
def is_audio_in_length_range(length):
|
391 |
+
return length > min_input_length and length < max_input_length
|
392 |
+
|
393 |
+
vectorized_datasets = vectorized_datasets.filter(
|
394 |
+
is_audio_in_length_range,
|
395 |
+
num_proc=num_workers,
|
396 |
+
input_columns=["input_length"],
|
397 |
+
)
|
398 |
+
|
399 |
+
# for large datasets it is advised to run the preprocessing on a
|
400 |
+
# single machine first with `args.preprocessing_only` since there will mostly likely
|
401 |
+
# be a timeout when running the script in distributed mode.
|
402 |
+
# In a second step `args.preprocessing_only` can then be set to `False` to load the
|
403 |
+
# cached dataset
|
404 |
+
if data_args.preprocessing_only:
|
405 |
+
cache = {k: v.cache_files for k, v in vectorized_datasets.items()}
|
406 |
+
logger.info(f"Data preprocessing finished. Files cached at {cache}.")
|
407 |
+
return
|
408 |
+
|
409 |
+
# 8. Load Metric
|
410 |
+
metric = load_metric("wer")
|
411 |
+
|
412 |
+
def compute_metrics(pred):
|
413 |
+
pred_ids = pred.predictions
|
414 |
+
|
415 |
+
pred.label_ids[pred.label_ids == -100] = tokenizer.pad_token_id
|
416 |
+
|
417 |
+
pred_str = tokenizer.batch_decode(pred_ids, skip_special_tokens=True)
|
418 |
+
# we do not want to group tokens when computing the metrics
|
419 |
+
label_str = tokenizer.batch_decode(pred.label_ids, skip_special_tokens=True)
|
420 |
+
|
421 |
+
wer = metric.compute(predictions=pred_str, references=label_str)
|
422 |
+
|
423 |
+
return {"wer": wer}
|
424 |
+
|
425 |
+
# 9. Create a single speech processor
|
426 |
+
if is_main_process(training_args.local_rank):
|
427 |
+
# save feature extractor, tokenizer and config
|
428 |
+
feature_extractor.save_pretrained(training_args.output_dir)
|
429 |
+
tokenizer.save_pretrained(training_args.output_dir)
|
430 |
+
config.save_pretrained(training_args.output_dir)
|
431 |
+
|
432 |
+
processor = AutoProcessor.from_pretrained(training_args.output_dir)
|
433 |
+
|
434 |
+
# 10. Define data collator
|
435 |
+
data_collator = DataCollatorSpeechSeq2SeqWithPadding(
|
436 |
+
processor=processor, decoder_start_token_id=model.config.decoder_start_token_id
|
437 |
+
)
|
438 |
+
|
439 |
+
decay_parameters = get_parameter_names(model, [torch.nn.LayerNorm])
|
440 |
+
decay_parameters = [name for name in decay_parameters if "bias" not in name]
|
441 |
+
optimizer_grouped_parameters = [
|
442 |
+
{
|
443 |
+
"params": [p for n, p in model.named_parameters() if n in decay_parameters],
|
444 |
+
"weight_decay": training_args.weight_decay,
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"params": [p for n, p in model.named_parameters() if n not in decay_parameters],
|
448 |
+
"weight_decay": 0.0,
|
449 |
+
},
|
450 |
+
]
|
451 |
+
|
452 |
+
optimizer = bnb.optim.Adam8bit(
|
453 |
+
params=optimizer_grouped_parameters,
|
454 |
+
lr=training_args.learning_rate,
|
455 |
+
betas=(training_args.adam_beta1, training_args.adam_beta2),
|
456 |
+
eps=training_args.adam_epsilon,
|
457 |
+
)
|
458 |
+
|
459 |
+
"""optimizer = Adafactor(
|
460 |
+
params=optimizer_grouped_parameters,
|
461 |
+
lr=training_args.learning_rate,
|
462 |
+
beta1=training_args.adam_beta1,
|
463 |
+
eps=training_args.adam_epsilon,
|
464 |
+
relative_step=False,
|
465 |
+
)"""
|
466 |
+
|
467 |
+
|
468 |
+
optimizers = (optimizer, None)
|
469 |
+
|
470 |
+
|
471 |
+
#11. Initialize Trainer
|
472 |
+
|
473 |
+
trainer = Seq2SeqTrainer(
|
474 |
+
model=model,
|
475 |
+
args=training_args,
|
476 |
+
train_dataset=vectorized_datasets["train"] if training_args.do_train else None,
|
477 |
+
eval_dataset=vectorized_datasets["eval"] if training_args.do_eval else None,
|
478 |
+
tokenizer=feature_extractor,
|
479 |
+
data_collator=data_collator,
|
480 |
+
compute_metrics=compute_metrics if training_args.predict_with_generate else None,
|
481 |
+
optimizers=optimizers,
|
482 |
+
)
|
483 |
+
|
484 |
+
# 12. Training
|
485 |
+
if training_args.do_train:
|
486 |
+
checkpoint = None
|
487 |
+
if training_args.resume_from_checkpoint is not None:
|
488 |
+
checkpoint = training_args.resume_from_checkpoint
|
489 |
+
elif last_checkpoint is not None:
|
490 |
+
checkpoint = last_checkpoint
|
491 |
+
train_result = trainer.train(resume_from_checkpoint=checkpoint)
|
492 |
+
trainer.save_model() # Saves the feature extractor too for easy upload
|
493 |
+
|
494 |
+
metrics = train_result.metrics
|
495 |
+
max_train_samples = (
|
496 |
+
data_args.max_train_samples
|
497 |
+
if data_args.max_train_samples is not None
|
498 |
+
else len(vectorized_datasets["train"])
|
499 |
+
)
|
500 |
+
metrics["train_samples"] = min(max_train_samples, len(vectorized_datasets["train"]))
|
501 |
+
trainer.log_metrics("train", metrics)
|
502 |
+
trainer.save_metrics("train", metrics)
|
503 |
+
trainer.save_state()
|
504 |
+
|
505 |
+
# 13. Evaluation
|
506 |
+
results = {}
|
507 |
+
if training_args.do_eval:
|
508 |
+
logger.info("*** Evaluate ***")
|
509 |
+
metrics = trainer.evaluate(
|
510 |
+
metric_key_prefix="eval", max_length=model.config.max_length, num_beams=model.config.num_beams
|
511 |
+
)
|
512 |
+
max_eval_samples = (
|
513 |
+
data_args.max_eval_samples if data_args.max_eval_samples is not None else len(vectorized_datasets["eval"])
|
514 |
+
)
|
515 |
+
metrics["eval_samples"] = min(max_eval_samples, len(vectorized_datasets["eval"]))
|
516 |
+
|
517 |
+
trainer.log_metrics("eval", metrics)
|
518 |
+
trainer.save_metrics("eval", metrics)
|
519 |
+
|
520 |
+
# 14. Write Training Stats
|
521 |
+
kwargs = {"finetuned_from": model_args.model_name_or_path, "tasks": "speech recognition"}
|
522 |
+
if data_args.dataset_name is not None:
|
523 |
+
kwargs["dataset_tags"] = data_args.dataset_name
|
524 |
+
if data_args.dataset_config_name is not None:
|
525 |
+
kwargs["dataset_args"] = data_args.dataset_config_name
|
526 |
+
kwargs["dataset"] = f"{data_args.dataset_name} {data_args.dataset_config_name}"
|
527 |
+
else:
|
528 |
+
kwargs["dataset"] = data_args.dataset_name
|
529 |
+
|
530 |
+
if training_args.push_to_hub:
|
531 |
+
trainer.push_to_hub(**kwargs)
|
532 |
+
else:
|
533 |
+
trainer.create_model_card(**kwargs)
|
534 |
+
|
535 |
+
return results
|
536 |
+
|
537 |
+
|
538 |
+
if __name__ == "__main__":
|
539 |
+
main()
|
runs/Mar10_21-08-09_sanchit--v100/1646946516.4121053/events.out.tfevents.1646946516.sanchit--v100.1040434.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:082322fbba064c4add0d46f619c710a1909fd2dac276569d0919120ab307b069
|
3 |
+
size 4973
|
runs/Mar10_21-08-09_sanchit--v100/events.out.tfevents.1646946516.sanchit--v100.1040434.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ef0a14d56e0d90b20191f8713147b42cca9dc405c25d98383f769a34d74214b
|
3 |
+
size 9899
|
runs/Mar10_21-09-23_sanchit--v100/1646946590.0115986/events.out.tfevents.1646946590.sanchit--v100.1043009.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6c190ca2d619b1821f8485caaf00456f555c8401582b324d6d197652e314af7b
|
3 |
+
size 4973
|
runs/Mar10_21-09-23_sanchit--v100/events.out.tfevents.1646946590.sanchit--v100.1043009.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4ad8fcd0f3ab5a26f83cfc43a9cf727a2a9a260b362f952c28736d3b965265ed
|
3 |
+
size 9593
|
runs/Mar10_21-20-12_sanchit--v100/1646947238.8154027/events.out.tfevents.1646947238.sanchit--v100.1065811.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:67042b29234691d5f587ac52aaa8fe1268630a6f93b70faf48f5042e0292107c
|
3 |
+
size 4973
|
runs/Mar10_21-20-12_sanchit--v100/events.out.tfevents.1646947238.sanchit--v100.1065811.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f8ecb528c2a7343952147f0e372a600fe80531bb2aad4cf7479369439fbba23d
|
3 |
+
size 245028
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "./", "tokenizer_class": "BartTokenizer"}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8eeaec852c625839280c4dc63b28753f4396524bd117a99fa66de2980131662a
|
3 |
+
size 3119
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
wandb/debug-internal.log
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
run-20220310_212039-1nikmzrh/logs/debug-internal.log
|
wandb/debug.log
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
run-20220310_212039-1nikmzrh/logs/debug.log
|
wandb/latest-run
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
run-20220310_212039-1nikmzrh
|
wandb/run-20220310_210836-1mziq47o/files/config.yaml
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
wandb_version: 1
|
2 |
+
|
3 |
+
_wandb:
|
4 |
+
desc: null
|
5 |
+
value:
|
6 |
+
cli_version: 0.12.10
|
7 |
+
framework: huggingface
|
8 |
+
huggingface_version: 4.17.0.dev0
|
9 |
+
is_jupyter_run: false
|
10 |
+
is_kaggle_kernel: false
|
11 |
+
python_version: 3.9.5
|
12 |
+
start_time: 1646946516
|
13 |
+
t:
|
14 |
+
1:
|
15 |
+
- 1
|
16 |
+
- 5
|
17 |
+
- 11
|
18 |
+
3:
|
19 |
+
- 13
|
20 |
+
4: 3.9.5
|
21 |
+
5: 0.12.10
|
22 |
+
6: 4.17.0.dev0
|
23 |
+
8:
|
24 |
+
- 5
|
wandb/run-20220310_210836-1mziq47o/files/output.log
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
|
3 |
+
0%| | 0/17840 [00:00<?, ?it/s][WARNING|modeling_utils.py:388] 2022-03-10 21:08:40,241 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
4 |
+
[WARNING|modeling_utils.py:388] 2022-03-10 21:08:42,264 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
5 |
+
[WARNING|modeling_utils.py:388] 2022-03-10 21:08:44,412 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
6 |
+
{'loss': 11.0815, 'learning_rate': 3.0000000000000004e-08, 'epoch': 0.0}
|
7 |
+
[WARNING|modeling_utils.py:388] 2022-03-10 21:08:46,393 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
8 |
+
|
9 |
+
0%| | 1/17840 [00:08<42:27:12, 8.57s/it][WARNING|modeling_utils.py:388] 2022-03-10 21:08:48,389 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
10 |
+
[WARNING|modeling_utils.py:388] 2022-03-10 21:08:50,170 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
11 |
+
{'loss': 11.0883, 'learning_rate': 6.000000000000001e-08, 'epoch': 0.0}
|
wandb/run-20220310_210836-1mziq47o/files/requirements.txt
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
absl-py==1.0.0
|
2 |
+
aiohttp==3.8.1
|
3 |
+
aiosignal==1.2.0
|
4 |
+
anyio==3.5.0
|
5 |
+
appdirs==1.4.4
|
6 |
+
argon2-cffi-bindings==21.2.0
|
7 |
+
argon2-cffi==21.3.0
|
8 |
+
asttokens==2.0.5
|
9 |
+
async-timeout==4.0.2
|
10 |
+
attrs==21.4.0
|
11 |
+
audioread==2.1.9
|
12 |
+
babel==2.9.1
|
13 |
+
backcall==0.2.0
|
14 |
+
bitsandbytes-cuda113==0.26.0
|
15 |
+
black==22.1.0
|
16 |
+
bleach==4.1.0
|
17 |
+
cachetools==5.0.0
|
18 |
+
certifi==2021.10.8
|
19 |
+
cffi==1.15.0
|
20 |
+
charset-normalizer==2.0.11
|
21 |
+
chex==0.1.0
|
22 |
+
click==8.0.3
|
23 |
+
clldutils==3.10.1
|
24 |
+
colorlog==6.6.0
|
25 |
+
csvw==1.11.0
|
26 |
+
cycler==0.11.0
|
27 |
+
datasets==1.18.3
|
28 |
+
debugpy==1.5.1
|
29 |
+
decorator==5.1.1
|
30 |
+
defusedxml==0.7.1
|
31 |
+
dill==0.3.4
|
32 |
+
dlinfo==1.2.1
|
33 |
+
dm-tree==0.1.6
|
34 |
+
docker-pycreds==0.4.0
|
35 |
+
entrypoints==0.4
|
36 |
+
executing==0.8.2
|
37 |
+
filelock==3.4.2
|
38 |
+
flatbuffers==2.0
|
39 |
+
flax==0.4.0
|
40 |
+
fonttools==4.29.1
|
41 |
+
frozenlist==1.3.0
|
42 |
+
fsspec==2022.1.0
|
43 |
+
gitdb==4.0.9
|
44 |
+
gitpython==3.1.27
|
45 |
+
google-auth-oauthlib==0.4.6
|
46 |
+
google-auth==2.6.0
|
47 |
+
grpcio==1.43.0
|
48 |
+
huggingface-hub==0.4.0
|
49 |
+
hypothesis==6.36.1
|
50 |
+
idna==3.3
|
51 |
+
importlib-metadata==4.10.1
|
52 |
+
ipykernel==6.8.0
|
53 |
+
ipython-genutils==0.2.0
|
54 |
+
ipython==8.0.1
|
55 |
+
ipywidgets==7.6.5
|
56 |
+
isodate==0.6.1
|
57 |
+
jax==0.2.28
|
58 |
+
jaxlib==0.1.76+cuda11.cudnn82
|
59 |
+
jedi==0.18.1
|
60 |
+
jinja2==3.0.3
|
61 |
+
jiwer==2.3.0
|
62 |
+
joblib==1.1.0
|
63 |
+
json5==0.9.6
|
64 |
+
jsonschema==4.4.0
|
65 |
+
jupyter-client==7.1.2
|
66 |
+
jupyter-console==6.4.0
|
67 |
+
jupyter-core==4.9.1
|
68 |
+
jupyter-server==1.13.5
|
69 |
+
jupyter==1.0.0
|
70 |
+
jupyterlab-pygments==0.1.2
|
71 |
+
jupyterlab-server==2.10.3
|
72 |
+
jupyterlab-widgets==1.0.2
|
73 |
+
jupyterlab==3.2.9
|
74 |
+
kiwisolver==1.3.2
|
75 |
+
librosa==0.8.1
|
76 |
+
llvmlite==0.38.0
|
77 |
+
markdown==3.3.6
|
78 |
+
markupsafe==2.0.1
|
79 |
+
matplotlib-inline==0.1.3
|
80 |
+
matplotlib==3.5.1
|
81 |
+
mistune==0.8.4
|
82 |
+
msgpack==1.0.3
|
83 |
+
multidict==6.0.2
|
84 |
+
multiprocess==0.70.12.2
|
85 |
+
mypy-extensions==0.4.3
|
86 |
+
nbclassic==0.3.5
|
87 |
+
nbclient==0.5.10
|
88 |
+
nbconvert==6.4.1
|
89 |
+
nbformat==5.1.3
|
90 |
+
nest-asyncio==1.5.4
|
91 |
+
notebook==6.4.8
|
92 |
+
numba==0.55.1
|
93 |
+
numpy==1.21.5
|
94 |
+
oauthlib==3.2.0
|
95 |
+
opt-einsum==3.3.0
|
96 |
+
optax==0.1.0
|
97 |
+
packaging==21.3
|
98 |
+
pandas==1.4.0
|
99 |
+
pandocfilters==1.5.0
|
100 |
+
parso==0.8.3
|
101 |
+
pathspec==0.9.0
|
102 |
+
pathtools==0.1.2
|
103 |
+
pexpect==4.8.0
|
104 |
+
phonemizer==3.0.1
|
105 |
+
pickleshare==0.7.5
|
106 |
+
pillow==9.0.0
|
107 |
+
pip==22.0.2
|
108 |
+
pkg-resources==0.0.0
|
109 |
+
platformdirs==2.4.1
|
110 |
+
pooch==1.6.0
|
111 |
+
prometheus-client==0.13.1
|
112 |
+
promise==2.3
|
113 |
+
prompt-toolkit==3.0.26
|
114 |
+
protobuf==3.19.4
|
115 |
+
psutil==5.9.0
|
116 |
+
ptyprocess==0.7.0
|
117 |
+
pure-eval==0.2.2
|
118 |
+
pyarrow==6.0.1
|
119 |
+
pyasn1-modules==0.2.8
|
120 |
+
pyasn1==0.4.8
|
121 |
+
pycparser==2.21
|
122 |
+
pyctcdecode==0.3.0
|
123 |
+
pygments==2.11.2
|
124 |
+
pygtrie==2.4.2
|
125 |
+
pyparsing==3.0.7
|
126 |
+
pyrsistent==0.18.1
|
127 |
+
python-dateutil==2.8.2
|
128 |
+
python-levenshtein==0.12.2
|
129 |
+
pytz==2021.3
|
130 |
+
pyyaml==6.0
|
131 |
+
pyzmq==22.3.0
|
132 |
+
qtconsole==5.2.2
|
133 |
+
qtpy==2.0.1
|
134 |
+
regex==2022.1.18
|
135 |
+
requests-oauthlib==1.3.1
|
136 |
+
requests==2.27.1
|
137 |
+
resampy==0.2.2
|
138 |
+
rfc3986==2.0.0
|
139 |
+
rsa==4.8
|
140 |
+
sacremoses==0.0.47
|
141 |
+
scikit-learn==1.0.2
|
142 |
+
scipy==1.7.3
|
143 |
+
segments==2.2.0
|
144 |
+
send2trash==1.8.0
|
145 |
+
sentry-sdk==1.5.6
|
146 |
+
setuptools==44.1.1
|
147 |
+
shortuuid==1.0.8
|
148 |
+
six==1.16.0
|
149 |
+
smmap==5.0.0
|
150 |
+
sniffio==1.2.0
|
151 |
+
sortedcontainers==2.4.0
|
152 |
+
soundfile==0.10.3.post1
|
153 |
+
stack-data==0.1.4
|
154 |
+
tabulate==0.8.9
|
155 |
+
tensorboard-data-server==0.6.1
|
156 |
+
tensorboard-plugin-wit==1.8.1
|
157 |
+
tensorboard==2.8.0
|
158 |
+
termcolor==1.1.0
|
159 |
+
terminado==0.13.1
|
160 |
+
testpath==0.5.0
|
161 |
+
threadpoolctl==3.1.0
|
162 |
+
tokenizers==0.11.4
|
163 |
+
tomli==2.0.0
|
164 |
+
toolz==0.11.2
|
165 |
+
torch==1.10.2+cu113
|
166 |
+
torchaudio==0.10.2+cu113
|
167 |
+
tornado==6.1
|
168 |
+
tqdm==4.62.3
|
169 |
+
traitlets==5.1.1
|
170 |
+
transformers==4.17.0.dev0
|
171 |
+
typing-extensions==3.10.0.2
|
172 |
+
uritemplate==4.1.1
|
173 |
+
urllib3==1.26.8
|
174 |
+
wandb==0.12.10
|
175 |
+
wcwidth==0.2.5
|
176 |
+
webencodings==0.5.1
|
177 |
+
websocket-client==1.2.3
|
178 |
+
werkzeug==2.0.2
|
179 |
+
wheel==0.37.1
|
180 |
+
widgetsnbextension==3.5.2
|
181 |
+
xxhash==2.0.2
|
182 |
+
yarl==1.7.2
|
183 |
+
yaspin==2.1.0
|
184 |
+
zipp==3.7.0
|
wandb/run-20220310_210836-1mziq47o/files/wandb-metadata.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"os": "Linux-5.11.0-1028-gcp-x86_64-with-glibc2.33",
|
3 |
+
"python": "3.9.5",
|
4 |
+
"heartbeatAt": "2022-03-10T21:08:37.861008",
|
5 |
+
"startedAt": "2022-03-10T21:08:36.772111",
|
6 |
+
"docker": null,
|
7 |
+
"gpu": "Tesla V100-SXM2-16GB",
|
8 |
+
"gpu_count": 2,
|
9 |
+
"cpu_count": 16,
|
10 |
+
"cuda": null,
|
11 |
+
"args": [
|
12 |
+
"--dataset_name=librispeech_asr",
|
13 |
+
"--model_name_or_path=./",
|
14 |
+
"--tokenizer_name=./",
|
15 |
+
"--dataset_config_name=clean",
|
16 |
+
"--train_split_name=train.100",
|
17 |
+
"--eval_split_name=validation",
|
18 |
+
"--output_dir=./",
|
19 |
+
"--preprocessing_num_workers=1",
|
20 |
+
"--length_column_name=input_length",
|
21 |
+
"--overwrite_output_dir",
|
22 |
+
"--num_train_epochs=20",
|
23 |
+
"--per_device_train_batch_size=8",
|
24 |
+
"--per_device_eval_batch_size=8",
|
25 |
+
"--gradient_accumulation_steps=4",
|
26 |
+
"--generation_max_length=40",
|
27 |
+
"--generation_num_beams=1",
|
28 |
+
"--learning_rate=3e-5",
|
29 |
+
"--warmup_steps=1000",
|
30 |
+
"--evaluation_strategy=steps",
|
31 |
+
"--text_column_name=text",
|
32 |
+
"--save_steps=1500",
|
33 |
+
"--eval_steps=1500",
|
34 |
+
"--logging_steps=1",
|
35 |
+
"--save_total_limit=1",
|
36 |
+
"--freeze_feature_encoder",
|
37 |
+
"--gradient_checkpointing",
|
38 |
+
"--fp16",
|
39 |
+
"--group_by_length",
|
40 |
+
"--predict_with_generate",
|
41 |
+
"--do_lower_case",
|
42 |
+
"--do_eval",
|
43 |
+
"--do_train",
|
44 |
+
"--push_to_hub",
|
45 |
+
"--use_auth_token"
|
46 |
+
],
|
47 |
+
"state": "running",
|
48 |
+
"program": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py",
|
49 |
+
"codePath": "run_speech_recognition_seq2seq.py",
|
50 |
+
"git": {
|
51 |
+
"remote": "https://huggingface.co/sanchit-gandhi/wav2vec2-2-rnd-2-layer-bart",
|
52 |
+
"commit": "4ea7f446103e7b0489f13a4de255ec90e6838d3b"
|
53 |
+
},
|
54 |
+
"email": "sanchit@huggingface.co",
|
55 |
+
"root": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart",
|
56 |
+
"host": "sanchit--v100",
|
57 |
+
"username": "sanchit_huggingface_co",
|
58 |
+
"executable": "/home/sanchit_huggingface_co/gcp/bin/python"
|
59 |
+
}
|
wandb/run-20220310_210836-1mziq47o/files/wandb-summary.json
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
{"train/loss": 11.0883, "train/learning_rate": 6.000000000000001e-08, "train/epoch": 0.0, "train/global_step": 2, "_runtime": 17, "_timestamp": 1646946533, "_step": 1}
|
wandb/run-20220310_210836-1mziq47o/logs/debug-internal.log
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
2022-03-10 21:08:37,687 INFO MainThread:1041406 [internal.py:wandb_internal():89] W&B internal server running at pid: 1041406, started at: 2022-03-10 21:08:37.687430
|
2 |
+
2022-03-10 21:08:37,689 DEBUG HandlerThread:1041406 [handler.py:handle_request():131] handle_request: check_version
|
3 |
+
2022-03-10 21:08:37,690 INFO WriterThread:1041406 [datastore.py:open_for_write():77] open: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/run-1mziq47o.wandb
|
4 |
+
2022-03-10 21:08:37,691 DEBUG SenderThread:1041406 [sender.py:send():235] send: header
|
5 |
+
2022-03-10 21:08:37,691 DEBUG SenderThread:1041406 [sender.py:send_request():249] send_request: check_version
|
6 |
+
2022-03-10 21:08:37,760 DEBUG SenderThread:1041406 [sender.py:send():235] send: run
|
7 |
+
2022-03-10 21:08:37,855 INFO SenderThread:1041406 [dir_watcher.py:__init__():169] watching files in: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files
|
8 |
+
2022-03-10 21:08:37,855 INFO SenderThread:1041406 [sender.py:_start_run_threads():809] run started: 1mziq47o with start time 1646946516
|
9 |
+
2022-03-10 21:08:37,855 DEBUG SenderThread:1041406 [sender.py:send():235] send: summary
|
10 |
+
2022-03-10 21:08:37,856 INFO SenderThread:1041406 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
11 |
+
2022-03-10 21:08:37,856 DEBUG HandlerThread:1041406 [handler.py:handle_request():131] handle_request: run_start
|
12 |
+
2022-03-10 21:08:37,860 DEBUG HandlerThread:1041406 [meta.py:__init__():36] meta init
|
13 |
+
2022-03-10 21:08:37,860 DEBUG HandlerThread:1041406 [meta.py:__init__():50] meta init done
|
14 |
+
2022-03-10 21:08:37,860 DEBUG HandlerThread:1041406 [meta.py:probe():210] probe
|
15 |
+
2022-03-10 21:08:37,867 DEBUG HandlerThread:1041406 [meta.py:_setup_git():200] setup git
|
16 |
+
2022-03-10 21:08:37,881 DEBUG HandlerThread:1041406 [meta.py:_setup_git():207] setup git done
|
17 |
+
2022-03-10 21:08:37,881 DEBUG HandlerThread:1041406 [meta.py:_save_pip():54] save pip
|
18 |
+
2022-03-10 21:08:37,882 DEBUG HandlerThread:1041406 [meta.py:_save_pip():68] save pip done
|
19 |
+
2022-03-10 21:08:37,882 DEBUG HandlerThread:1041406 [meta.py:probe():248] probe done
|
20 |
+
2022-03-10 21:08:37,967 DEBUG SenderThread:1041406 [sender.py:send():235] send: files
|
21 |
+
2022-03-10 21:08:37,967 INFO SenderThread:1041406 [sender.py:_save_file():944] saving file wandb-metadata.json with policy now
|
22 |
+
2022-03-10 21:08:37,974 DEBUG HandlerThread:1041406 [handler.py:handle_request():131] handle_request: stop_status
|
23 |
+
2022-03-10 21:08:37,974 DEBUG SenderThread:1041406 [sender.py:send_request():249] send_request: stop_status
|
24 |
+
2022-03-10 21:08:38,010 DEBUG SenderThread:1041406 [sender.py:send():235] send: config
|
25 |
+
2022-03-10 21:08:38,011 DEBUG SenderThread:1041406 [sender.py:send():235] send: metric
|
26 |
+
2022-03-10 21:08:38,011 DEBUG SenderThread:1041406 [sender.py:send():235] send: metric
|
27 |
+
2022-03-10 21:08:38,011 WARNING SenderThread:1041406 [sender.py:send_metric():902] Seen metric with glob (shouldnt happen)
|
28 |
+
2022-03-10 21:08:38,286 INFO Thread-11 :1041406 [upload_job.py:push():137] Uploaded file /tmp/tmp62yhv5cewandb/5rx4zxrq-wandb-metadata.json
|
29 |
+
2022-03-10 21:08:38,857 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/wandb-metadata.json
|
30 |
+
2022-03-10 21:08:38,858 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/wandb-summary.json
|
31 |
+
2022-03-10 21:08:38,858 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/requirements.txt
|
32 |
+
2022-03-10 21:08:38,858 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
33 |
+
2022-03-10 21:08:40,856 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
34 |
+
2022-03-10 21:08:42,857 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
35 |
+
2022-03-10 21:08:44,858 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
36 |
+
2022-03-10 21:08:46,561 DEBUG SenderThread:1041406 [sender.py:send():235] send: metric
|
37 |
+
2022-03-10 21:08:46,561 DEBUG SenderThread:1041406 [sender.py:send():235] send: metric
|
38 |
+
2022-03-10 21:08:46,561 DEBUG SenderThread:1041406 [sender.py:send():235] send: metric
|
39 |
+
2022-03-10 21:08:46,561 DEBUG SenderThread:1041406 [sender.py:send():235] send: history
|
40 |
+
2022-03-10 21:08:46,561 DEBUG SenderThread:1041406 [sender.py:send():235] send: summary
|
41 |
+
2022-03-10 21:08:46,562 INFO SenderThread:1041406 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
42 |
+
2022-03-10 21:08:46,859 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/wandb-summary.json
|
43 |
+
2022-03-10 21:08:46,859 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
44 |
+
2022-03-10 21:08:48,859 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
45 |
+
2022-03-10 21:08:50,860 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
46 |
+
2022-03-10 21:08:52,861 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
47 |
+
2022-03-10 21:08:53,241 DEBUG HandlerThread:1041406 [handler.py:handle_request():131] handle_request: stop_status
|
48 |
+
2022-03-10 21:08:53,241 DEBUG SenderThread:1041406 [sender.py:send_request():249] send_request: stop_status
|
49 |
+
2022-03-10 21:08:53,753 DEBUG SenderThread:1041406 [sender.py:send():235] send: history
|
50 |
+
2022-03-10 21:08:53,754 DEBUG SenderThread:1041406 [sender.py:send():235] send: summary
|
51 |
+
2022-03-10 21:08:53,754 INFO SenderThread:1041406 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
52 |
+
2022-03-10 21:08:53,861 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/wandb-summary.json
|
53 |
+
2022-03-10 21:08:54,861 INFO Thread-8 :1041406 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/files/output.log
|
54 |
+
2022-03-10 21:08:55,247 WARNING MainThread:1041406 [internal.py:wandb_internal():152] Internal process interrupt: 1
|
55 |
+
2022-03-10 21:08:55,393 WARNING MainThread:1041406 [internal.py:wandb_internal():152] Internal process interrupt: 2
|
56 |
+
2022-03-10 21:08:55,393 ERROR MainThread:1041406 [internal.py:wandb_internal():155] Internal process interrupted.
|
57 |
+
2022-03-10 21:08:55,736 INFO MainThread:1041406 [internal.py:handle_exit():79] Internal process exited
|
wandb/run-20220310_210836-1mziq47o/logs/debug.log
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/.config/wandb/settings
|
2 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/settings
|
3 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_setup.py:_flush():75] Loading settings from environment variables: {}
|
4 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_setup.py:_flush():75] Inferring run settings from compute environment: {'program_relpath': 'run_speech_recognition_seq2seq.py', 'program': '/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py'}
|
5 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_init.py:_log_setup():386] Logging user logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/logs/debug.log
|
6 |
+
2022-03-10 21:08:36,773 INFO MainThread:1040434 [wandb_init.py:_log_setup():387] Logging internal logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210836-1mziq47o/logs/debug-internal.log
|
7 |
+
2022-03-10 21:08:36,774 INFO MainThread:1040434 [wandb_init.py:init():420] calling init triggers
|
8 |
+
2022-03-10 21:08:36,774 INFO MainThread:1040434 [wandb_init.py:init():425] wandb.init called with sweep_config: {}
|
9 |
+
config: {}
|
10 |
+
2022-03-10 21:08:36,774 INFO MainThread:1040434 [wandb_init.py:init():471] starting backend
|
11 |
+
2022-03-10 21:08:36,774 INFO MainThread:1040434 [backend.py:_multiprocessing_setup():99] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
|
12 |
+
2022-03-10 21:08:36,836 INFO MainThread:1040434 [backend.py:ensure_launched():219] starting backend process...
|
13 |
+
2022-03-10 21:08:36,896 INFO MainThread:1040434 [backend.py:ensure_launched():224] started backend process with pid: 1041406
|
14 |
+
2022-03-10 21:08:36,898 INFO MainThread:1040434 [wandb_init.py:init():480] backend started and connected
|
15 |
+
2022-03-10 21:08:36,907 INFO MainThread:1040434 [wandb_init.py:init():550] updated telemetry
|
16 |
+
2022-03-10 21:08:37,051 INFO MainThread:1040434 [wandb_init.py:init():581] communicating current version
|
17 |
+
2022-03-10 21:08:37,758 INFO MainThread:1040434 [wandb_init.py:init():586] got version response upgrade_message: "wandb version 0.12.11 is available! To upgrade, please run:\n $ pip install wandb --upgrade"
|
18 |
+
|
19 |
+
2022-03-10 21:08:37,759 INFO MainThread:1040434 [wandb_init.py:init():596] communicating run to backend with 30 second timeout
|
20 |
+
2022-03-10 21:08:37,856 INFO MainThread:1040434 [wandb_init.py:init():624] starting run threads in backend
|
21 |
+
2022-03-10 21:08:37,971 INFO MainThread:1040434 [wandb_run.py:_console_start():1827] atexit reg
|
22 |
+
2022-03-10 21:08:37,972 INFO MainThread:1040434 [wandb_run.py:_redirect():1701] redirect: SettingsConsole.REDIRECT
|
23 |
+
2022-03-10 21:08:37,972 INFO MainThread:1040434 [wandb_run.py:_redirect():1706] Redirecting console.
|
24 |
+
2022-03-10 21:08:37,974 INFO MainThread:1040434 [wandb_run.py:_redirect():1762] Redirects installed.
|
25 |
+
2022-03-10 21:08:37,974 INFO MainThread:1040434 [wandb_init.py:init():651] run started, returning control to user process
|
26 |
+
2022-03-10 21:08:37,976 INFO MainThread:1040434 [wandb_run.py:_config_callback():966] config_cb None None {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'torch.float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'is_encoder_decoder': True, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 50, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['SpeechEncoderDecoderModel'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 0, 'task_specific_params': None, 'problem_type': None, '_name_or_path': './', 'transformers_version': None, 'decoder': {'vocab_size': 50265, 'max_position_embeddings': 1024, 'd_model': 1024, 'encoder_ffn_dim': 4096, 'encoder_layers': 12, 'encoder_attention_heads': 16, 'decoder_ffn_dim': 4096, 'decoder_layers': 12, 'decoder_attention_heads': 16, 'dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'activation_function': 'gelu', 'init_std': 0.02, 'encoder_layerdrop': 0.0, 'decoder_layerdrop': 0.0, 'classifier_dropout': 0.0, 'use_cache': False, 'num_hidden_layers': 2, 'scale_embedding': False, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': True, 'cross_attention_hidden_size': None, 'add_cross_attention': True, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': True, 'num_beams': 4, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 3, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': 0, 'forced_eos_token_id': 2, 'remove_invalid_values': False, 'architectures': ['BartForCausalLM'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1', 2: 'LABEL_2'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1, 'LABEL_2': 2}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 0, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 2, 'task_specific_params': {'summarization': {'length_penalty': 1.0, 'max_length': 128, 'min_length': 12, 'num_beams': 4}, 'summarization_cnn': {'length_penalty': 2.0, 'max_length': 142, 'min_length': 56, 'num_beams': 4}, 'summarization_xsum': {'length_penalty': 1.0, 'max_length': 62, 'min_length': 11, 'num_beams': 6}}, 'problem_type': None, '_name_or_path': 'decoder', 'transformers_version': '4.17.0.dev0', 'add_bias_logits': False, 'add_final_layer_norm': False, 'classif_dropout': 0.0, 'gradient_checkpointing': False, 'normalize_before': False, 'model_type': 'bart'}, 'encoder': {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': None, 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['Wav2Vec2ForPreTraining'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'facebook/wav2vec2-large-lv60', 'transformers_version': '4.17.0.dev0', 'feat_extract_dropout': 0.0, 'gradient_checkpointing': False, 'hidden_dropout_prob': 0.0, 'num_feat_extract_layers': 7, 'hidden_size': 1024, 'feat_extract_norm': 'layer', 'feat_extract_activation': 'gelu', 'conv_dim': [512, 512, 512, 512, 512, 512, 512], 'conv_stride': [5, 2, 2, 2, 2, 2, 2], 'conv_kernel': [10, 3, 3, 3, 3, 2, 2], 'conv_bias': True, 'num_conv_pos_embeddings': 128, 'num_conv_pos_embedding_groups': 16, 'num_hidden_layers': 24, 'intermediate_size': 4096, 'hidden_act': 'gelu', 'num_attention_heads': 16, 'hidden_dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'feat_proj_dropout': 0.0, 'final_dropout': 0.0, 'layerdrop': 0.0, 'layer_norm_eps': 1e-05, 'initializer_range': 0.02, 'vocab_size': 32, 'do_stable_layer_norm': True, 'use_weighted_layer_sum': False, 'apply_spec_augment': False, 'mask_time_prob': 0.0, 'mask_time_length': 10, 'mask_time_min_masks': 2, 'mask_feature_prob': 0.0, 'mask_feature_length': 10, 'mask_feature_min_masks': 0, 'num_codevectors_per_group': 320, 'num_codevector_groups': 2, 'contrastive_logits_temperature': 0.1, 'feat_quantizer_dropout': 0.0, 'num_negatives': 100, 'codevector_dim': 768, 'proj_codevector_dim': 768, 'diversity_loss_weight': 0.1, 'ctc_loss_reduction': 'sum', 'ctc_zero_infinity': False, 'add_adapter': True, 'adapter_kernel_size': 3, 'adapter_stride': 2, 'num_adapter_layers': 3, 'output_hidden_size': 1024, 'classifier_proj_size': 256, 'tdnn_dim': [512, 512, 512, 512, 1500], 'tdnn_kernel': [5, 3, 3, 1, 1], 'tdnn_dilation': [1, 2, 3, 1, 1], 'xvector_output_dim': 512, 'model_type': 'wav2vec2'}, 'model_type': 'speech-encoder-decoder', 'processor_class': 'Wav2Vec2Processor', 'use_cache': False, 'output_dir': './', 'overwrite_output_dir': True, 'do_train': True, 'do_eval': True, 'do_predict': False, 'evaluation_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 8, 'per_device_eval_batch_size': 8, 'per_gpu_train_batch_size': 'None', 'per_gpu_eval_batch_size': 'None', 'gradient_accumulation_steps': 4, 'eval_accumulation_steps': 'None', 'learning_rate': 3e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 20.0, 'max_steps': -1, 'lr_scheduler_type': 'linear', 'warmup_ratio': 0.0, 'warmup_steps': 1000, 'log_level': -1, 'log_level_replica': -1, 'log_on_each_node': True, 'logging_dir': './runs/Mar10_21-08-09_sanchit--v100', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 1500, 'save_total_limit': 1, 'save_on_each_node': False, 'no_cuda': False, 'seed': 42, 'bf16': False, 'fp16': True, 'fp16_opt_level': 'O1', 'half_precision_backend': 'amp', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': 'None', 'local_rank': -1, 'xpu_backend': 'None', 'tpu_num_cores': 'None', 'tpu_metrics_debug': False, 'debug': '[]', 'dataloader_drop_last': False, 'eval_steps': 1500, 'dataloader_num_workers': 0, 'past_index': -1, 'run_name': './', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': 'None', 'load_best_model_at_end': False, 'metric_for_best_model': 'None', 'greater_is_better': 'None', 'ignore_data_skip': False, 'sharded_ddp': '[]', 'deepspeed': 'None', 'label_smoothing_factor': 0.0, 'optim': 'adamw_hf', 'adafactor': False, 'group_by_length': True, 'length_column_name': 'input_length', 'report_to': "['tensorboard', 'wandb']", 'ddp_find_unused_parameters': 'None', 'ddp_bucket_cap_mb': 'None', 'dataloader_pin_memory': True, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': True, 'resume_from_checkpoint': 'None', 'hub_model_id': 'None', 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'gradient_checkpointing': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': 'None', 'push_to_hub_organization': 'None', 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', '_n_gpu': 1, 'mp_parameters': '', 'sortish_sampler': False, 'predict_with_generate': True, 'generation_max_length': 40, 'generation_num_beams': 1, 'train_batch_size': 8, 'eval_batch_size': 8}
|
27 |
+
2022-03-10 21:08:37,979 INFO MainThread:1040434 [wandb_watch.py:watch():43] Watching
|
28 |
+
2022-03-10 21:08:55,321 INFO MainThread:1040434 [wandb_run.py:_atexit_cleanup():1797] got exitcode: 255
|
29 |
+
2022-03-10 21:08:55,322 INFO MainThread:1040434 [wandb_run.py:_restore():1769] restore
|
wandb/run-20220310_210836-1mziq47o/run-1mziq47o.wandb
ADDED
Binary file (9.52 kB). View file
|
wandb/run-20220310_210950-19sg19ou/files/config.yaml
ADDED
@@ -0,0 +1,728 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
wandb_version: 1
|
2 |
+
|
3 |
+
_n_gpu:
|
4 |
+
desc: null
|
5 |
+
value: 1
|
6 |
+
_name_or_path:
|
7 |
+
desc: null
|
8 |
+
value: ./
|
9 |
+
_wandb:
|
10 |
+
desc: null
|
11 |
+
value:
|
12 |
+
cli_version: 0.12.10
|
13 |
+
framework: huggingface
|
14 |
+
huggingface_version: 4.17.0.dev0
|
15 |
+
is_jupyter_run: false
|
16 |
+
is_kaggle_kernel: false
|
17 |
+
m:
|
18 |
+
- 1: train/global_step
|
19 |
+
6:
|
20 |
+
- 3
|
21 |
+
python_version: 3.9.5
|
22 |
+
start_time: 1646946590
|
23 |
+
t:
|
24 |
+
1:
|
25 |
+
- 1
|
26 |
+
- 5
|
27 |
+
- 11
|
28 |
+
2:
|
29 |
+
- 1
|
30 |
+
- 5
|
31 |
+
- 11
|
32 |
+
3:
|
33 |
+
- 1
|
34 |
+
- 7
|
35 |
+
- 13
|
36 |
+
4: 3.9.5
|
37 |
+
5: 0.12.10
|
38 |
+
6: 4.17.0.dev0
|
39 |
+
8:
|
40 |
+
- 5
|
41 |
+
adafactor:
|
42 |
+
desc: null
|
43 |
+
value: false
|
44 |
+
adam_beta1:
|
45 |
+
desc: null
|
46 |
+
value: 0.9
|
47 |
+
adam_beta2:
|
48 |
+
desc: null
|
49 |
+
value: 0.999
|
50 |
+
adam_epsilon:
|
51 |
+
desc: null
|
52 |
+
value: 1.0e-08
|
53 |
+
add_cross_attention:
|
54 |
+
desc: null
|
55 |
+
value: false
|
56 |
+
architectures:
|
57 |
+
desc: null
|
58 |
+
value:
|
59 |
+
- SpeechEncoderDecoderModel
|
60 |
+
bad_words_ids:
|
61 |
+
desc: null
|
62 |
+
value: null
|
63 |
+
bf16:
|
64 |
+
desc: null
|
65 |
+
value: false
|
66 |
+
bf16_full_eval:
|
67 |
+
desc: null
|
68 |
+
value: false
|
69 |
+
bos_token_id:
|
70 |
+
desc: null
|
71 |
+
value: null
|
72 |
+
chunk_size_feed_forward:
|
73 |
+
desc: null
|
74 |
+
value: 0
|
75 |
+
cross_attention_hidden_size:
|
76 |
+
desc: null
|
77 |
+
value: null
|
78 |
+
dataloader_drop_last:
|
79 |
+
desc: null
|
80 |
+
value: false
|
81 |
+
dataloader_num_workers:
|
82 |
+
desc: null
|
83 |
+
value: 0
|
84 |
+
dataloader_pin_memory:
|
85 |
+
desc: null
|
86 |
+
value: true
|
87 |
+
ddp_bucket_cap_mb:
|
88 |
+
desc: null
|
89 |
+
value: None
|
90 |
+
ddp_find_unused_parameters:
|
91 |
+
desc: null
|
92 |
+
value: None
|
93 |
+
debug:
|
94 |
+
desc: null
|
95 |
+
value: '[]'
|
96 |
+
decoder:
|
97 |
+
desc: null
|
98 |
+
value:
|
99 |
+
_name_or_path: decoder
|
100 |
+
activation_dropout: 0.0
|
101 |
+
activation_function: gelu
|
102 |
+
add_bias_logits: false
|
103 |
+
add_cross_attention: true
|
104 |
+
add_final_layer_norm: false
|
105 |
+
architectures:
|
106 |
+
- BartForCausalLM
|
107 |
+
attention_dropout: 0.0
|
108 |
+
bad_words_ids: null
|
109 |
+
bos_token_id: 0
|
110 |
+
chunk_size_feed_forward: 0
|
111 |
+
classif_dropout: 0.0
|
112 |
+
classifier_dropout: 0.0
|
113 |
+
cross_attention_hidden_size: null
|
114 |
+
d_model: 1024
|
115 |
+
decoder_attention_heads: 16
|
116 |
+
decoder_ffn_dim: 4096
|
117 |
+
decoder_layerdrop: 0.0
|
118 |
+
decoder_layers: 12
|
119 |
+
decoder_start_token_id: 2
|
120 |
+
diversity_penalty: 0.0
|
121 |
+
do_sample: false
|
122 |
+
dropout: 0.0
|
123 |
+
early_stopping: true
|
124 |
+
encoder_attention_heads: 16
|
125 |
+
encoder_ffn_dim: 4096
|
126 |
+
encoder_layerdrop: 0.0
|
127 |
+
encoder_layers: 12
|
128 |
+
encoder_no_repeat_ngram_size: 0
|
129 |
+
eos_token_id: 2
|
130 |
+
finetuning_task: null
|
131 |
+
forced_bos_token_id: 0
|
132 |
+
forced_eos_token_id: 2
|
133 |
+
gradient_checkpointing: false
|
134 |
+
id2label:
|
135 |
+
'0': LABEL_0
|
136 |
+
'1': LABEL_1
|
137 |
+
'2': LABEL_2
|
138 |
+
init_std: 0.02
|
139 |
+
is_decoder: true
|
140 |
+
is_encoder_decoder: false
|
141 |
+
label2id:
|
142 |
+
LABEL_0: 0
|
143 |
+
LABEL_1: 1
|
144 |
+
LABEL_2: 2
|
145 |
+
length_penalty: 1.0
|
146 |
+
max_length: 20
|
147 |
+
max_position_embeddings: 1024
|
148 |
+
min_length: 0
|
149 |
+
model_type: bart
|
150 |
+
no_repeat_ngram_size: 3
|
151 |
+
normalize_before: false
|
152 |
+
num_beam_groups: 1
|
153 |
+
num_beams: 4
|
154 |
+
num_hidden_layers: 2
|
155 |
+
num_return_sequences: 1
|
156 |
+
output_attentions: false
|
157 |
+
output_hidden_states: false
|
158 |
+
output_scores: false
|
159 |
+
pad_token_id: 1
|
160 |
+
prefix: null
|
161 |
+
problem_type: null
|
162 |
+
pruned_heads: {}
|
163 |
+
remove_invalid_values: false
|
164 |
+
repetition_penalty: 1.0
|
165 |
+
return_dict: true
|
166 |
+
return_dict_in_generate: false
|
167 |
+
scale_embedding: false
|
168 |
+
sep_token_id: null
|
169 |
+
task_specific_params:
|
170 |
+
summarization:
|
171 |
+
length_penalty: 1.0
|
172 |
+
max_length: 128
|
173 |
+
min_length: 12
|
174 |
+
num_beams: 4
|
175 |
+
summarization_cnn:
|
176 |
+
length_penalty: 2.0
|
177 |
+
max_length: 142
|
178 |
+
min_length: 56
|
179 |
+
num_beams: 4
|
180 |
+
summarization_xsum:
|
181 |
+
length_penalty: 1.0
|
182 |
+
max_length: 62
|
183 |
+
min_length: 11
|
184 |
+
num_beams: 6
|
185 |
+
temperature: 1.0
|
186 |
+
tie_encoder_decoder: false
|
187 |
+
tie_word_embeddings: true
|
188 |
+
tokenizer_class: null
|
189 |
+
top_k: 50
|
190 |
+
top_p: 1.0
|
191 |
+
torch_dtype: float32
|
192 |
+
torchscript: false
|
193 |
+
transformers_version: 4.17.0.dev0
|
194 |
+
use_bfloat16: false
|
195 |
+
use_cache: false
|
196 |
+
vocab_size: 50265
|
197 |
+
decoder_start_token_id:
|
198 |
+
desc: null
|
199 |
+
value: 0
|
200 |
+
deepspeed:
|
201 |
+
desc: null
|
202 |
+
value: None
|
203 |
+
disable_tqdm:
|
204 |
+
desc: null
|
205 |
+
value: false
|
206 |
+
diversity_penalty:
|
207 |
+
desc: null
|
208 |
+
value: 0.0
|
209 |
+
do_eval:
|
210 |
+
desc: null
|
211 |
+
value: true
|
212 |
+
do_predict:
|
213 |
+
desc: null
|
214 |
+
value: false
|
215 |
+
do_sample:
|
216 |
+
desc: null
|
217 |
+
value: false
|
218 |
+
do_train:
|
219 |
+
desc: null
|
220 |
+
value: true
|
221 |
+
early_stopping:
|
222 |
+
desc: null
|
223 |
+
value: false
|
224 |
+
encoder:
|
225 |
+
desc: null
|
226 |
+
value:
|
227 |
+
_name_or_path: facebook/wav2vec2-large-lv60
|
228 |
+
activation_dropout: 0.0
|
229 |
+
adapter_kernel_size: 3
|
230 |
+
adapter_stride: 2
|
231 |
+
add_adapter: true
|
232 |
+
add_cross_attention: false
|
233 |
+
apply_spec_augment: false
|
234 |
+
architectures:
|
235 |
+
- Wav2Vec2ForPreTraining
|
236 |
+
attention_dropout: 0.0
|
237 |
+
bad_words_ids: null
|
238 |
+
bos_token_id: 1
|
239 |
+
chunk_size_feed_forward: 0
|
240 |
+
classifier_proj_size: 256
|
241 |
+
codevector_dim: 768
|
242 |
+
contrastive_logits_temperature: 0.1
|
243 |
+
conv_bias: true
|
244 |
+
conv_dim:
|
245 |
+
- 512
|
246 |
+
- 512
|
247 |
+
- 512
|
248 |
+
- 512
|
249 |
+
- 512
|
250 |
+
- 512
|
251 |
+
- 512
|
252 |
+
conv_kernel:
|
253 |
+
- 10
|
254 |
+
- 3
|
255 |
+
- 3
|
256 |
+
- 3
|
257 |
+
- 3
|
258 |
+
- 2
|
259 |
+
- 2
|
260 |
+
conv_stride:
|
261 |
+
- 5
|
262 |
+
- 2
|
263 |
+
- 2
|
264 |
+
- 2
|
265 |
+
- 2
|
266 |
+
- 2
|
267 |
+
- 2
|
268 |
+
cross_attention_hidden_size: null
|
269 |
+
ctc_loss_reduction: sum
|
270 |
+
ctc_zero_infinity: false
|
271 |
+
decoder_start_token_id: null
|
272 |
+
diversity_loss_weight: 0.1
|
273 |
+
diversity_penalty: 0.0
|
274 |
+
do_sample: false
|
275 |
+
do_stable_layer_norm: true
|
276 |
+
early_stopping: false
|
277 |
+
encoder_no_repeat_ngram_size: 0
|
278 |
+
eos_token_id: 2
|
279 |
+
feat_extract_activation: gelu
|
280 |
+
feat_extract_dropout: 0.0
|
281 |
+
feat_extract_norm: layer
|
282 |
+
feat_proj_dropout: 0.0
|
283 |
+
feat_quantizer_dropout: 0.0
|
284 |
+
final_dropout: 0.0
|
285 |
+
finetuning_task: null
|
286 |
+
forced_bos_token_id: null
|
287 |
+
forced_eos_token_id: null
|
288 |
+
gradient_checkpointing: false
|
289 |
+
hidden_act: gelu
|
290 |
+
hidden_dropout: 0.0
|
291 |
+
hidden_dropout_prob: 0.0
|
292 |
+
hidden_size: 1024
|
293 |
+
id2label:
|
294 |
+
'0': LABEL_0
|
295 |
+
'1': LABEL_1
|
296 |
+
initializer_range: 0.02
|
297 |
+
intermediate_size: 4096
|
298 |
+
is_decoder: false
|
299 |
+
is_encoder_decoder: false
|
300 |
+
label2id:
|
301 |
+
LABEL_0: 0
|
302 |
+
LABEL_1: 1
|
303 |
+
layer_norm_eps: 1.0e-05
|
304 |
+
layerdrop: 0.0
|
305 |
+
length_penalty: 1.0
|
306 |
+
mask_feature_length: 10
|
307 |
+
mask_feature_min_masks: 0
|
308 |
+
mask_feature_prob: 0.0
|
309 |
+
mask_time_length: 10
|
310 |
+
mask_time_min_masks: 2
|
311 |
+
mask_time_prob: 0.0
|
312 |
+
max_length: 20
|
313 |
+
min_length: 0
|
314 |
+
model_type: wav2vec2
|
315 |
+
no_repeat_ngram_size: 0
|
316 |
+
num_adapter_layers: 3
|
317 |
+
num_attention_heads: 16
|
318 |
+
num_beam_groups: 1
|
319 |
+
num_beams: 1
|
320 |
+
num_codevector_groups: 2
|
321 |
+
num_codevectors_per_group: 320
|
322 |
+
num_conv_pos_embedding_groups: 16
|
323 |
+
num_conv_pos_embeddings: 128
|
324 |
+
num_feat_extract_layers: 7
|
325 |
+
num_hidden_layers: 24
|
326 |
+
num_negatives: 100
|
327 |
+
num_return_sequences: 1
|
328 |
+
output_attentions: false
|
329 |
+
output_hidden_size: 1024
|
330 |
+
output_hidden_states: false
|
331 |
+
output_scores: false
|
332 |
+
pad_token_id: 0
|
333 |
+
prefix: null
|
334 |
+
problem_type: null
|
335 |
+
proj_codevector_dim: 768
|
336 |
+
pruned_heads: {}
|
337 |
+
remove_invalid_values: false
|
338 |
+
repetition_penalty: 1.0
|
339 |
+
return_dict: true
|
340 |
+
return_dict_in_generate: false
|
341 |
+
sep_token_id: null
|
342 |
+
task_specific_params: null
|
343 |
+
tdnn_dilation:
|
344 |
+
- 1
|
345 |
+
- 2
|
346 |
+
- 3
|
347 |
+
- 1
|
348 |
+
- 1
|
349 |
+
tdnn_dim:
|
350 |
+
- 512
|
351 |
+
- 512
|
352 |
+
- 512
|
353 |
+
- 512
|
354 |
+
- 1500
|
355 |
+
tdnn_kernel:
|
356 |
+
- 5
|
357 |
+
- 3
|
358 |
+
- 3
|
359 |
+
- 1
|
360 |
+
- 1
|
361 |
+
temperature: 1.0
|
362 |
+
tie_encoder_decoder: false
|
363 |
+
tie_word_embeddings: true
|
364 |
+
tokenizer_class: null
|
365 |
+
top_k: 50
|
366 |
+
top_p: 1.0
|
367 |
+
torch_dtype: null
|
368 |
+
torchscript: false
|
369 |
+
transformers_version: 4.17.0.dev0
|
370 |
+
use_bfloat16: false
|
371 |
+
use_weighted_layer_sum: false
|
372 |
+
vocab_size: 32
|
373 |
+
xvector_output_dim: 512
|
374 |
+
encoder_no_repeat_ngram_size:
|
375 |
+
desc: null
|
376 |
+
value: 0
|
377 |
+
eos_token_id:
|
378 |
+
desc: null
|
379 |
+
value: 2
|
380 |
+
eval_accumulation_steps:
|
381 |
+
desc: null
|
382 |
+
value: None
|
383 |
+
eval_batch_size:
|
384 |
+
desc: null
|
385 |
+
value: 16
|
386 |
+
eval_steps:
|
387 |
+
desc: null
|
388 |
+
value: 1500
|
389 |
+
evaluation_strategy:
|
390 |
+
desc: null
|
391 |
+
value: steps
|
392 |
+
finetuning_task:
|
393 |
+
desc: null
|
394 |
+
value: null
|
395 |
+
forced_bos_token_id:
|
396 |
+
desc: null
|
397 |
+
value: null
|
398 |
+
forced_eos_token_id:
|
399 |
+
desc: null
|
400 |
+
value: null
|
401 |
+
fp16:
|
402 |
+
desc: null
|
403 |
+
value: true
|
404 |
+
fp16_backend:
|
405 |
+
desc: null
|
406 |
+
value: auto
|
407 |
+
fp16_full_eval:
|
408 |
+
desc: null
|
409 |
+
value: false
|
410 |
+
fp16_opt_level:
|
411 |
+
desc: null
|
412 |
+
value: O1
|
413 |
+
generation_max_length:
|
414 |
+
desc: null
|
415 |
+
value: 40
|
416 |
+
generation_num_beams:
|
417 |
+
desc: null
|
418 |
+
value: 1
|
419 |
+
gradient_accumulation_steps:
|
420 |
+
desc: null
|
421 |
+
value: 8
|
422 |
+
gradient_checkpointing:
|
423 |
+
desc: null
|
424 |
+
value: true
|
425 |
+
greater_is_better:
|
426 |
+
desc: null
|
427 |
+
value: None
|
428 |
+
group_by_length:
|
429 |
+
desc: null
|
430 |
+
value: true
|
431 |
+
half_precision_backend:
|
432 |
+
desc: null
|
433 |
+
value: amp
|
434 |
+
hub_model_id:
|
435 |
+
desc: null
|
436 |
+
value: None
|
437 |
+
hub_strategy:
|
438 |
+
desc: null
|
439 |
+
value: every_save
|
440 |
+
hub_token:
|
441 |
+
desc: null
|
442 |
+
value: <HUB_TOKEN>
|
443 |
+
id2label:
|
444 |
+
desc: null
|
445 |
+
value:
|
446 |
+
'0': LABEL_0
|
447 |
+
'1': LABEL_1
|
448 |
+
ignore_data_skip:
|
449 |
+
desc: null
|
450 |
+
value: false
|
451 |
+
is_decoder:
|
452 |
+
desc: null
|
453 |
+
value: false
|
454 |
+
is_encoder_decoder:
|
455 |
+
desc: null
|
456 |
+
value: true
|
457 |
+
label2id:
|
458 |
+
desc: null
|
459 |
+
value:
|
460 |
+
LABEL_0: 0
|
461 |
+
LABEL_1: 1
|
462 |
+
label_names:
|
463 |
+
desc: null
|
464 |
+
value: None
|
465 |
+
label_smoothing_factor:
|
466 |
+
desc: null
|
467 |
+
value: 0.0
|
468 |
+
learning_rate:
|
469 |
+
desc: null
|
470 |
+
value: 3.0e-05
|
471 |
+
length_column_name:
|
472 |
+
desc: null
|
473 |
+
value: input_length
|
474 |
+
length_penalty:
|
475 |
+
desc: null
|
476 |
+
value: 1.0
|
477 |
+
load_best_model_at_end:
|
478 |
+
desc: null
|
479 |
+
value: false
|
480 |
+
local_rank:
|
481 |
+
desc: null
|
482 |
+
value: -1
|
483 |
+
log_level:
|
484 |
+
desc: null
|
485 |
+
value: -1
|
486 |
+
log_level_replica:
|
487 |
+
desc: null
|
488 |
+
value: -1
|
489 |
+
log_on_each_node:
|
490 |
+
desc: null
|
491 |
+
value: true
|
492 |
+
logging_dir:
|
493 |
+
desc: null
|
494 |
+
value: ./runs/Mar10_21-09-23_sanchit--v100
|
495 |
+
logging_first_step:
|
496 |
+
desc: null
|
497 |
+
value: false
|
498 |
+
logging_nan_inf_filter:
|
499 |
+
desc: null
|
500 |
+
value: true
|
501 |
+
logging_steps:
|
502 |
+
desc: null
|
503 |
+
value: 1
|
504 |
+
logging_strategy:
|
505 |
+
desc: null
|
506 |
+
value: steps
|
507 |
+
lr_scheduler_type:
|
508 |
+
desc: null
|
509 |
+
value: linear
|
510 |
+
max_grad_norm:
|
511 |
+
desc: null
|
512 |
+
value: 1.0
|
513 |
+
max_length:
|
514 |
+
desc: null
|
515 |
+
value: 50
|
516 |
+
max_steps:
|
517 |
+
desc: null
|
518 |
+
value: -1
|
519 |
+
metric_for_best_model:
|
520 |
+
desc: null
|
521 |
+
value: None
|
522 |
+
min_length:
|
523 |
+
desc: null
|
524 |
+
value: 0
|
525 |
+
model_type:
|
526 |
+
desc: null
|
527 |
+
value: speech-encoder-decoder
|
528 |
+
mp_parameters:
|
529 |
+
desc: null
|
530 |
+
value: ''
|
531 |
+
no_cuda:
|
532 |
+
desc: null
|
533 |
+
value: false
|
534 |
+
no_repeat_ngram_size:
|
535 |
+
desc: null
|
536 |
+
value: 0
|
537 |
+
num_beam_groups:
|
538 |
+
desc: null
|
539 |
+
value: 1
|
540 |
+
num_beams:
|
541 |
+
desc: null
|
542 |
+
value: 1
|
543 |
+
num_return_sequences:
|
544 |
+
desc: null
|
545 |
+
value: 1
|
546 |
+
num_train_epochs:
|
547 |
+
desc: null
|
548 |
+
value: 20.0
|
549 |
+
optim:
|
550 |
+
desc: null
|
551 |
+
value: adamw_hf
|
552 |
+
output_attentions:
|
553 |
+
desc: null
|
554 |
+
value: false
|
555 |
+
output_dir:
|
556 |
+
desc: null
|
557 |
+
value: ./
|
558 |
+
output_hidden_states:
|
559 |
+
desc: null
|
560 |
+
value: false
|
561 |
+
output_scores:
|
562 |
+
desc: null
|
563 |
+
value: false
|
564 |
+
overwrite_output_dir:
|
565 |
+
desc: null
|
566 |
+
value: true
|
567 |
+
pad_token_id:
|
568 |
+
desc: null
|
569 |
+
value: 1
|
570 |
+
past_index:
|
571 |
+
desc: null
|
572 |
+
value: -1
|
573 |
+
per_device_eval_batch_size:
|
574 |
+
desc: null
|
575 |
+
value: 16
|
576 |
+
per_device_train_batch_size:
|
577 |
+
desc: null
|
578 |
+
value: 16
|
579 |
+
per_gpu_eval_batch_size:
|
580 |
+
desc: null
|
581 |
+
value: None
|
582 |
+
per_gpu_train_batch_size:
|
583 |
+
desc: null
|
584 |
+
value: None
|
585 |
+
predict_with_generate:
|
586 |
+
desc: null
|
587 |
+
value: true
|
588 |
+
prediction_loss_only:
|
589 |
+
desc: null
|
590 |
+
value: false
|
591 |
+
prefix:
|
592 |
+
desc: null
|
593 |
+
value: null
|
594 |
+
problem_type:
|
595 |
+
desc: null
|
596 |
+
value: null
|
597 |
+
processor_class:
|
598 |
+
desc: null
|
599 |
+
value: Wav2Vec2Processor
|
600 |
+
pruned_heads:
|
601 |
+
desc: null
|
602 |
+
value: {}
|
603 |
+
push_to_hub:
|
604 |
+
desc: null
|
605 |
+
value: true
|
606 |
+
push_to_hub_model_id:
|
607 |
+
desc: null
|
608 |
+
value: None
|
609 |
+
push_to_hub_organization:
|
610 |
+
desc: null
|
611 |
+
value: None
|
612 |
+
push_to_hub_token:
|
613 |
+
desc: null
|
614 |
+
value: <PUSH_TO_HUB_TOKEN>
|
615 |
+
remove_invalid_values:
|
616 |
+
desc: null
|
617 |
+
value: false
|
618 |
+
remove_unused_columns:
|
619 |
+
desc: null
|
620 |
+
value: true
|
621 |
+
repetition_penalty:
|
622 |
+
desc: null
|
623 |
+
value: 1.0
|
624 |
+
report_to:
|
625 |
+
desc: null
|
626 |
+
value: '[''tensorboard'', ''wandb'']'
|
627 |
+
resume_from_checkpoint:
|
628 |
+
desc: null
|
629 |
+
value: None
|
630 |
+
return_dict:
|
631 |
+
desc: null
|
632 |
+
value: true
|
633 |
+
return_dict_in_generate:
|
634 |
+
desc: null
|
635 |
+
value: false
|
636 |
+
run_name:
|
637 |
+
desc: null
|
638 |
+
value: ./
|
639 |
+
save_on_each_node:
|
640 |
+
desc: null
|
641 |
+
value: false
|
642 |
+
save_steps:
|
643 |
+
desc: null
|
644 |
+
value: 1500
|
645 |
+
save_strategy:
|
646 |
+
desc: null
|
647 |
+
value: steps
|
648 |
+
save_total_limit:
|
649 |
+
desc: null
|
650 |
+
value: 1
|
651 |
+
seed:
|
652 |
+
desc: null
|
653 |
+
value: 42
|
654 |
+
sep_token_id:
|
655 |
+
desc: null
|
656 |
+
value: null
|
657 |
+
sharded_ddp:
|
658 |
+
desc: null
|
659 |
+
value: '[]'
|
660 |
+
skip_memory_metrics:
|
661 |
+
desc: null
|
662 |
+
value: true
|
663 |
+
sortish_sampler:
|
664 |
+
desc: null
|
665 |
+
value: false
|
666 |
+
task_specific_params:
|
667 |
+
desc: null
|
668 |
+
value: null
|
669 |
+
temperature:
|
670 |
+
desc: null
|
671 |
+
value: 1.0
|
672 |
+
tf32:
|
673 |
+
desc: null
|
674 |
+
value: None
|
675 |
+
tie_encoder_decoder:
|
676 |
+
desc: null
|
677 |
+
value: false
|
678 |
+
tie_word_embeddings:
|
679 |
+
desc: null
|
680 |
+
value: false
|
681 |
+
tokenizer_class:
|
682 |
+
desc: null
|
683 |
+
value: null
|
684 |
+
top_k:
|
685 |
+
desc: null
|
686 |
+
value: 50
|
687 |
+
top_p:
|
688 |
+
desc: null
|
689 |
+
value: 1.0
|
690 |
+
torch_dtype:
|
691 |
+
desc: null
|
692 |
+
value: torch.float32
|
693 |
+
torchscript:
|
694 |
+
desc: null
|
695 |
+
value: false
|
696 |
+
tpu_metrics_debug:
|
697 |
+
desc: null
|
698 |
+
value: false
|
699 |
+
tpu_num_cores:
|
700 |
+
desc: null
|
701 |
+
value: None
|
702 |
+
train_batch_size:
|
703 |
+
desc: null
|
704 |
+
value: 16
|
705 |
+
transformers_version:
|
706 |
+
desc: null
|
707 |
+
value: null
|
708 |
+
use_bfloat16:
|
709 |
+
desc: null
|
710 |
+
value: false
|
711 |
+
use_cache:
|
712 |
+
desc: null
|
713 |
+
value: false
|
714 |
+
use_legacy_prediction_loop:
|
715 |
+
desc: null
|
716 |
+
value: false
|
717 |
+
warmup_ratio:
|
718 |
+
desc: null
|
719 |
+
value: 0.0
|
720 |
+
warmup_steps:
|
721 |
+
desc: null
|
722 |
+
value: 1000
|
723 |
+
weight_decay:
|
724 |
+
desc: null
|
725 |
+
value: 0.0
|
726 |
+
xpu_backend:
|
727 |
+
desc: null
|
728 |
+
value: None
|
wandb/run-20220310_210950-19sg19ou/files/output.log
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
|
3 |
+
0%| | 0/4460 [00:00<?, ?it/s][WARNING|modeling_utils.py:388] 2022-03-10 21:09:55,515 >> Could not estimate the number of tokens of the input, floating-point operations will not be computed
|
4 |
+
Traceback (most recent call last):
|
5 |
+
File "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py", line 539, in <module>
|
6 |
+
main()
|
7 |
+
File "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py", line 491, in main
|
8 |
+
train_result = trainer.train(resume_from_checkpoint=checkpoint)
|
9 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/trainer.py", line 1384, in train
|
10 |
+
tr_loss_step = self.training_step(model, inputs)
|
11 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/trainer.py", line 1959, in training_step
|
12 |
+
loss = self.compute_loss(model, inputs)
|
13 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/trainer.py", line 1991, in compute_loss
|
14 |
+
outputs = model(**inputs)
|
15 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
|
16 |
+
return forward_call(*input, **kwargs)
|
17 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/models/speech_encoder_decoder/modeling_speech_encoder_decoder.py", line 503, in forward
|
18 |
+
encoder_outputs = self.encoder(
|
19 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
|
20 |
+
return forward_call(*input, **kwargs)
|
21 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/models/wav2vec2/modeling_wav2vec2.py", line 1346, in forward
|
22 |
+
extract_features = self.feature_extractor(input_values)
|
23 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
|
24 |
+
return forward_call(*input, **kwargs)
|
25 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/models/wav2vec2/modeling_wav2vec2.py", line 514, in forward
|
26 |
+
hidden_states = conv_layer(hidden_states)
|
27 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
|
28 |
+
return forward_call(*input, **kwargs)
|
29 |
+
File "/home/sanchit_huggingface_co/transformers/src/transformers/models/wav2vec2/modeling_wav2vec2.py", line 389, in forward
|
30 |
+
hidden_states = self.layer_norm(hidden_states)
|
31 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
|
32 |
+
return forward_call(*input, **kwargs)
|
33 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/modules/normalization.py", line 189, in forward
|
34 |
+
return F.layer_norm(
|
35 |
+
File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/torch/nn/functional.py", line 2347, in layer_norm
|
36 |
+
return torch.layer_norm(input, normalized_shape, weight, bias, eps, torch.backends.cudnn.enabled)
|
37 |
+
RuntimeError: CUDA out of memory. Tried to allocate 1.65 GiB (GPU 0; 15.78 GiB total capacity; 8.57 GiB already allocated; 1.58 GiB free; 12.60 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
|
wandb/run-20220310_210950-19sg19ou/files/requirements.txt
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
absl-py==1.0.0
|
2 |
+
aiohttp==3.8.1
|
3 |
+
aiosignal==1.2.0
|
4 |
+
anyio==3.5.0
|
5 |
+
appdirs==1.4.4
|
6 |
+
argon2-cffi-bindings==21.2.0
|
7 |
+
argon2-cffi==21.3.0
|
8 |
+
asttokens==2.0.5
|
9 |
+
async-timeout==4.0.2
|
10 |
+
attrs==21.4.0
|
11 |
+
audioread==2.1.9
|
12 |
+
babel==2.9.1
|
13 |
+
backcall==0.2.0
|
14 |
+
bitsandbytes-cuda113==0.26.0
|
15 |
+
black==22.1.0
|
16 |
+
bleach==4.1.0
|
17 |
+
cachetools==5.0.0
|
18 |
+
certifi==2021.10.8
|
19 |
+
cffi==1.15.0
|
20 |
+
charset-normalizer==2.0.11
|
21 |
+
chex==0.1.0
|
22 |
+
click==8.0.3
|
23 |
+
clldutils==3.10.1
|
24 |
+
colorlog==6.6.0
|
25 |
+
csvw==1.11.0
|
26 |
+
cycler==0.11.0
|
27 |
+
datasets==1.18.3
|
28 |
+
debugpy==1.5.1
|
29 |
+
decorator==5.1.1
|
30 |
+
defusedxml==0.7.1
|
31 |
+
dill==0.3.4
|
32 |
+
dlinfo==1.2.1
|
33 |
+
dm-tree==0.1.6
|
34 |
+
docker-pycreds==0.4.0
|
35 |
+
entrypoints==0.4
|
36 |
+
executing==0.8.2
|
37 |
+
filelock==3.4.2
|
38 |
+
flatbuffers==2.0
|
39 |
+
flax==0.4.0
|
40 |
+
fonttools==4.29.1
|
41 |
+
frozenlist==1.3.0
|
42 |
+
fsspec==2022.1.0
|
43 |
+
gitdb==4.0.9
|
44 |
+
gitpython==3.1.27
|
45 |
+
google-auth-oauthlib==0.4.6
|
46 |
+
google-auth==2.6.0
|
47 |
+
grpcio==1.43.0
|
48 |
+
huggingface-hub==0.4.0
|
49 |
+
hypothesis==6.36.1
|
50 |
+
idna==3.3
|
51 |
+
importlib-metadata==4.10.1
|
52 |
+
ipykernel==6.8.0
|
53 |
+
ipython-genutils==0.2.0
|
54 |
+
ipython==8.0.1
|
55 |
+
ipywidgets==7.6.5
|
56 |
+
isodate==0.6.1
|
57 |
+
jax==0.2.28
|
58 |
+
jaxlib==0.1.76+cuda11.cudnn82
|
59 |
+
jedi==0.18.1
|
60 |
+
jinja2==3.0.3
|
61 |
+
jiwer==2.3.0
|
62 |
+
joblib==1.1.0
|
63 |
+
json5==0.9.6
|
64 |
+
jsonschema==4.4.0
|
65 |
+
jupyter-client==7.1.2
|
66 |
+
jupyter-console==6.4.0
|
67 |
+
jupyter-core==4.9.1
|
68 |
+
jupyter-server==1.13.5
|
69 |
+
jupyter==1.0.0
|
70 |
+
jupyterlab-pygments==0.1.2
|
71 |
+
jupyterlab-server==2.10.3
|
72 |
+
jupyterlab-widgets==1.0.2
|
73 |
+
jupyterlab==3.2.9
|
74 |
+
kiwisolver==1.3.2
|
75 |
+
librosa==0.8.1
|
76 |
+
llvmlite==0.38.0
|
77 |
+
markdown==3.3.6
|
78 |
+
markupsafe==2.0.1
|
79 |
+
matplotlib-inline==0.1.3
|
80 |
+
matplotlib==3.5.1
|
81 |
+
mistune==0.8.4
|
82 |
+
msgpack==1.0.3
|
83 |
+
multidict==6.0.2
|
84 |
+
multiprocess==0.70.12.2
|
85 |
+
mypy-extensions==0.4.3
|
86 |
+
nbclassic==0.3.5
|
87 |
+
nbclient==0.5.10
|
88 |
+
nbconvert==6.4.1
|
89 |
+
nbformat==5.1.3
|
90 |
+
nest-asyncio==1.5.4
|
91 |
+
notebook==6.4.8
|
92 |
+
numba==0.55.1
|
93 |
+
numpy==1.21.5
|
94 |
+
oauthlib==3.2.0
|
95 |
+
opt-einsum==3.3.0
|
96 |
+
optax==0.1.0
|
97 |
+
packaging==21.3
|
98 |
+
pandas==1.4.0
|
99 |
+
pandocfilters==1.5.0
|
100 |
+
parso==0.8.3
|
101 |
+
pathspec==0.9.0
|
102 |
+
pathtools==0.1.2
|
103 |
+
pexpect==4.8.0
|
104 |
+
phonemizer==3.0.1
|
105 |
+
pickleshare==0.7.5
|
106 |
+
pillow==9.0.0
|
107 |
+
pip==22.0.2
|
108 |
+
pkg-resources==0.0.0
|
109 |
+
platformdirs==2.4.1
|
110 |
+
pooch==1.6.0
|
111 |
+
prometheus-client==0.13.1
|
112 |
+
promise==2.3
|
113 |
+
prompt-toolkit==3.0.26
|
114 |
+
protobuf==3.19.4
|
115 |
+
psutil==5.9.0
|
116 |
+
ptyprocess==0.7.0
|
117 |
+
pure-eval==0.2.2
|
118 |
+
pyarrow==6.0.1
|
119 |
+
pyasn1-modules==0.2.8
|
120 |
+
pyasn1==0.4.8
|
121 |
+
pycparser==2.21
|
122 |
+
pyctcdecode==0.3.0
|
123 |
+
pygments==2.11.2
|
124 |
+
pygtrie==2.4.2
|
125 |
+
pyparsing==3.0.7
|
126 |
+
pyrsistent==0.18.1
|
127 |
+
python-dateutil==2.8.2
|
128 |
+
python-levenshtein==0.12.2
|
129 |
+
pytz==2021.3
|
130 |
+
pyyaml==6.0
|
131 |
+
pyzmq==22.3.0
|
132 |
+
qtconsole==5.2.2
|
133 |
+
qtpy==2.0.1
|
134 |
+
regex==2022.1.18
|
135 |
+
requests-oauthlib==1.3.1
|
136 |
+
requests==2.27.1
|
137 |
+
resampy==0.2.2
|
138 |
+
rfc3986==2.0.0
|
139 |
+
rsa==4.8
|
140 |
+
sacremoses==0.0.47
|
141 |
+
scikit-learn==1.0.2
|
142 |
+
scipy==1.7.3
|
143 |
+
segments==2.2.0
|
144 |
+
send2trash==1.8.0
|
145 |
+
sentry-sdk==1.5.6
|
146 |
+
setuptools==44.1.1
|
147 |
+
shortuuid==1.0.8
|
148 |
+
six==1.16.0
|
149 |
+
smmap==5.0.0
|
150 |
+
sniffio==1.2.0
|
151 |
+
sortedcontainers==2.4.0
|
152 |
+
soundfile==0.10.3.post1
|
153 |
+
stack-data==0.1.4
|
154 |
+
tabulate==0.8.9
|
155 |
+
tensorboard-data-server==0.6.1
|
156 |
+
tensorboard-plugin-wit==1.8.1
|
157 |
+
tensorboard==2.8.0
|
158 |
+
termcolor==1.1.0
|
159 |
+
terminado==0.13.1
|
160 |
+
testpath==0.5.0
|
161 |
+
threadpoolctl==3.1.0
|
162 |
+
tokenizers==0.11.4
|
163 |
+
tomli==2.0.0
|
164 |
+
toolz==0.11.2
|
165 |
+
torch==1.10.2+cu113
|
166 |
+
torchaudio==0.10.2+cu113
|
167 |
+
tornado==6.1
|
168 |
+
tqdm==4.62.3
|
169 |
+
traitlets==5.1.1
|
170 |
+
transformers==4.17.0.dev0
|
171 |
+
typing-extensions==3.10.0.2
|
172 |
+
uritemplate==4.1.1
|
173 |
+
urllib3==1.26.8
|
174 |
+
wandb==0.12.10
|
175 |
+
wcwidth==0.2.5
|
176 |
+
webencodings==0.5.1
|
177 |
+
websocket-client==1.2.3
|
178 |
+
werkzeug==2.0.2
|
179 |
+
wheel==0.37.1
|
180 |
+
widgetsnbextension==3.5.2
|
181 |
+
xxhash==2.0.2
|
182 |
+
yarl==1.7.2
|
183 |
+
yaspin==2.1.0
|
184 |
+
zipp==3.7.0
|
wandb/run-20220310_210950-19sg19ou/files/wandb-metadata.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"os": "Linux-5.11.0-1028-gcp-x86_64-with-glibc2.33",
|
3 |
+
"python": "3.9.5",
|
4 |
+
"heartbeatAt": "2022-03-10T21:09:51.403094",
|
5 |
+
"startedAt": "2022-03-10T21:09:50.346010",
|
6 |
+
"docker": null,
|
7 |
+
"gpu": "Tesla V100-SXM2-16GB",
|
8 |
+
"gpu_count": 2,
|
9 |
+
"cpu_count": 16,
|
10 |
+
"cuda": null,
|
11 |
+
"args": [
|
12 |
+
"--dataset_name=librispeech_asr",
|
13 |
+
"--model_name_or_path=./",
|
14 |
+
"--tokenizer_name=./",
|
15 |
+
"--dataset_config_name=clean",
|
16 |
+
"--train_split_name=train.100",
|
17 |
+
"--eval_split_name=validation",
|
18 |
+
"--output_dir=./",
|
19 |
+
"--preprocessing_num_workers=1",
|
20 |
+
"--length_column_name=input_length",
|
21 |
+
"--overwrite_output_dir",
|
22 |
+
"--num_train_epochs=20",
|
23 |
+
"--per_device_train_batch_size=16",
|
24 |
+
"--per_device_eval_batch_size=16",
|
25 |
+
"--gradient_accumulation_steps=8",
|
26 |
+
"--generation_max_length=40",
|
27 |
+
"--generation_num_beams=1",
|
28 |
+
"--learning_rate=3e-5",
|
29 |
+
"--warmup_steps=1000",
|
30 |
+
"--evaluation_strategy=steps",
|
31 |
+
"--text_column_name=text",
|
32 |
+
"--save_steps=1500",
|
33 |
+
"--eval_steps=1500",
|
34 |
+
"--logging_steps=1",
|
35 |
+
"--save_total_limit=1",
|
36 |
+
"--freeze_feature_encoder",
|
37 |
+
"--gradient_checkpointing",
|
38 |
+
"--fp16",
|
39 |
+
"--group_by_length",
|
40 |
+
"--predict_with_generate",
|
41 |
+
"--do_lower_case",
|
42 |
+
"--do_eval",
|
43 |
+
"--do_train",
|
44 |
+
"--push_to_hub",
|
45 |
+
"--use_auth_token"
|
46 |
+
],
|
47 |
+
"state": "running",
|
48 |
+
"program": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py",
|
49 |
+
"codePath": "run_speech_recognition_seq2seq.py",
|
50 |
+
"git": {
|
51 |
+
"remote": "https://huggingface.co/sanchit-gandhi/wav2vec2-2-rnd-2-layer-bart",
|
52 |
+
"commit": "4ea7f446103e7b0489f13a4de255ec90e6838d3b"
|
53 |
+
},
|
54 |
+
"email": "sanchit@huggingface.co",
|
55 |
+
"root": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart",
|
56 |
+
"host": "sanchit--v100",
|
57 |
+
"username": "sanchit_huggingface_co",
|
58 |
+
"executable": "/home/sanchit_huggingface_co/gcp/bin/python"
|
59 |
+
}
|
wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
{"_wandb": {"runtime": 8}}
|
wandb/run-20220310_210950-19sg19ou/logs/debug-internal.log
ADDED
@@ -0,0 +1,137 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
2022-03-10 21:09:51,233 INFO MainThread:1044076 [internal.py:wandb_internal():89] W&B internal server running at pid: 1044076, started at: 2022-03-10 21:09:51.233570
|
2 |
+
2022-03-10 21:09:51,235 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: check_version
|
3 |
+
2022-03-10 21:09:51,235 INFO WriterThread:1044076 [datastore.py:open_for_write():77] open: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/run-19sg19ou.wandb
|
4 |
+
2022-03-10 21:09:51,237 DEBUG SenderThread:1044076 [sender.py:send():235] send: header
|
5 |
+
2022-03-10 21:09:51,237 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: check_version
|
6 |
+
2022-03-10 21:09:51,306 DEBUG SenderThread:1044076 [sender.py:send():235] send: run
|
7 |
+
2022-03-10 21:09:51,397 INFO SenderThread:1044076 [dir_watcher.py:__init__():169] watching files in: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files
|
8 |
+
2022-03-10 21:09:51,397 INFO SenderThread:1044076 [sender.py:_start_run_threads():809] run started: 19sg19ou with start time 1646946590
|
9 |
+
2022-03-10 21:09:51,398 DEBUG SenderThread:1044076 [sender.py:send():235] send: summary
|
10 |
+
2022-03-10 21:09:51,398 INFO SenderThread:1044076 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
11 |
+
2022-03-10 21:09:51,398 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: run_start
|
12 |
+
2022-03-10 21:09:51,402 DEBUG HandlerThread:1044076 [meta.py:__init__():36] meta init
|
13 |
+
2022-03-10 21:09:51,403 DEBUG HandlerThread:1044076 [meta.py:__init__():50] meta init done
|
14 |
+
2022-03-10 21:09:51,403 DEBUG HandlerThread:1044076 [meta.py:probe():210] probe
|
15 |
+
2022-03-10 21:09:51,410 DEBUG HandlerThread:1044076 [meta.py:_setup_git():200] setup git
|
16 |
+
2022-03-10 21:09:51,425 DEBUG HandlerThread:1044076 [meta.py:_setup_git():207] setup git done
|
17 |
+
2022-03-10 21:09:51,426 DEBUG HandlerThread:1044076 [meta.py:_save_pip():54] save pip
|
18 |
+
2022-03-10 21:09:51,426 DEBUG HandlerThread:1044076 [meta.py:_save_pip():68] save pip done
|
19 |
+
2022-03-10 21:09:51,426 DEBUG HandlerThread:1044076 [meta.py:probe():248] probe done
|
20 |
+
2022-03-10 21:09:51,505 DEBUG SenderThread:1044076 [sender.py:send():235] send: files
|
21 |
+
2022-03-10 21:09:51,505 INFO SenderThread:1044076 [sender.py:_save_file():944] saving file wandb-metadata.json with policy now
|
22 |
+
2022-03-10 21:09:51,509 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: stop_status
|
23 |
+
2022-03-10 21:09:51,510 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: stop_status
|
24 |
+
2022-03-10 21:09:51,544 DEBUG SenderThread:1044076 [sender.py:send():235] send: config
|
25 |
+
2022-03-10 21:09:51,545 DEBUG SenderThread:1044076 [sender.py:send():235] send: metric
|
26 |
+
2022-03-10 21:09:51,546 DEBUG SenderThread:1044076 [sender.py:send():235] send: metric
|
27 |
+
2022-03-10 21:09:51,546 WARNING SenderThread:1044076 [sender.py:send_metric():902] Seen metric with glob (shouldnt happen)
|
28 |
+
2022-03-10 21:09:51,790 INFO Thread-11 :1044076 [upload_job.py:push():137] Uploaded file /tmp/tmp0u4vh57hwandb/2s9kjsb8-wandb-metadata.json
|
29 |
+
2022-03-10 21:09:52,400 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json
|
30 |
+
2022-03-10 21:09:52,401 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/requirements.txt
|
31 |
+
2022-03-10 21:09:52,401 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log
|
32 |
+
2022-03-10 21:09:52,401 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_created():217] file/dir created: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-metadata.json
|
33 |
+
2022-03-10 21:09:54,398 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log
|
34 |
+
2022-03-10 21:09:58,400 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log
|
35 |
+
2022-03-10 21:09:59,980 DEBUG SenderThread:1044076 [sender.py:send():235] send: telemetry
|
36 |
+
2022-03-10 21:09:59,980 DEBUG SenderThread:1044076 [sender.py:send():235] send: exit
|
37 |
+
2022-03-10 21:09:59,980 INFO SenderThread:1044076 [sender.py:send_exit():371] handling exit code: 1
|
38 |
+
2022-03-10 21:09:59,981 INFO SenderThread:1044076 [sender.py:send_exit():373] handling runtime: 8
|
39 |
+
2022-03-10 21:09:59,981 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
40 |
+
2022-03-10 21:09:59,981 INFO SenderThread:1044076 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
41 |
+
2022-03-10 21:09:59,982 INFO SenderThread:1044076 [sender.py:send_exit():379] send defer
|
42 |
+
2022-03-10 21:09:59,982 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
43 |
+
2022-03-10 21:09:59,982 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
44 |
+
2022-03-10 21:09:59,982 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 0
|
45 |
+
2022-03-10 21:09:59,983 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
46 |
+
2022-03-10 21:09:59,983 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 0
|
47 |
+
2022-03-10 21:09:59,983 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 1
|
48 |
+
2022-03-10 21:09:59,983 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
49 |
+
2022-03-10 21:09:59,983 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 1
|
50 |
+
2022-03-10 21:10:00,043 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
51 |
+
2022-03-10 21:10:00,043 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 1
|
52 |
+
2022-03-10 21:10:00,043 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 2
|
53 |
+
2022-03-10 21:10:00,043 DEBUG SenderThread:1044076 [sender.py:send():235] send: stats
|
54 |
+
2022-03-10 21:10:00,044 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
55 |
+
2022-03-10 21:10:00,044 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 2
|
56 |
+
2022-03-10 21:10:00,044 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
57 |
+
2022-03-10 21:10:00,044 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 2
|
58 |
+
2022-03-10 21:10:00,044 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 3
|
59 |
+
2022-03-10 21:10:00,044 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
60 |
+
2022-03-10 21:10:00,045 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 3
|
61 |
+
2022-03-10 21:10:00,045 DEBUG SenderThread:1044076 [sender.py:send():235] send: summary
|
62 |
+
2022-03-10 21:10:00,045 INFO SenderThread:1044076 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
|
63 |
+
2022-03-10 21:10:00,045 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
64 |
+
2022-03-10 21:10:00,045 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 3
|
65 |
+
2022-03-10 21:10:00,045 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 4
|
66 |
+
2022-03-10 21:10:00,046 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
67 |
+
2022-03-10 21:10:00,046 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 4
|
68 |
+
2022-03-10 21:10:00,046 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
69 |
+
2022-03-10 21:10:00,046 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 4
|
70 |
+
2022-03-10 21:10:00,083 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
71 |
+
2022-03-10 21:10:00,172 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 5
|
72 |
+
2022-03-10 21:10:00,172 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
73 |
+
2022-03-10 21:10:00,173 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
74 |
+
2022-03-10 21:10:00,173 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 5
|
75 |
+
2022-03-10 21:10:00,173 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
76 |
+
2022-03-10 21:10:00,173 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 5
|
77 |
+
2022-03-10 21:10:00,173 INFO SenderThread:1044076 [dir_watcher.py:finish():283] shutting down directory watcher
|
78 |
+
2022-03-10 21:10:00,274 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
79 |
+
2022-03-10 21:10:00,401 INFO Thread-8 :1044076 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json
|
80 |
+
2022-03-10 21:10:00,401 INFO SenderThread:1044076 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/config.yaml
|
81 |
+
2022-03-10 21:10:00,401 INFO SenderThread:1044076 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log
|
82 |
+
2022-03-10 21:10:00,402 INFO SenderThread:1044076 [dir_watcher.py:finish():313] scan: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files
|
83 |
+
2022-03-10 21:10:00,402 INFO SenderThread:1044076 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-metadata.json wandb-metadata.json
|
84 |
+
2022-03-10 21:10:00,402 INFO SenderThread:1044076 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log output.log
|
85 |
+
2022-03-10 21:10:00,402 INFO SenderThread:1044076 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json wandb-summary.json
|
86 |
+
2022-03-10 21:10:00,402 INFO SenderThread:1044076 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/requirements.txt requirements.txt
|
87 |
+
2022-03-10 21:10:00,405 INFO SenderThread:1044076 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/config.yaml config.yaml
|
88 |
+
2022-03-10 21:10:00,408 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 6
|
89 |
+
2022-03-10 21:10:00,408 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
90 |
+
2022-03-10 21:10:00,414 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
91 |
+
2022-03-10 21:10:00,414 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 6
|
92 |
+
2022-03-10 21:10:00,415 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
93 |
+
2022-03-10 21:10:00,415 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 6
|
94 |
+
2022-03-10 21:10:00,415 INFO SenderThread:1044076 [file_pusher.py:finish():177] shutting down file pusher
|
95 |
+
2022-03-10 21:10:00,510 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
96 |
+
2022-03-10 21:10:00,510 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
97 |
+
2022-03-10 21:10:00,612 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
98 |
+
2022-03-10 21:10:00,612 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
99 |
+
2022-03-10 21:10:00,664 INFO Thread-12 :1044076 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/output.log
|
100 |
+
2022-03-10 21:10:00,672 INFO Thread-15 :1044076 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/config.yaml
|
101 |
+
2022-03-10 21:10:00,688 INFO Thread-13 :1044076 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/wandb-summary.json
|
102 |
+
2022-03-10 21:10:00,699 INFO Thread-14 :1044076 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/files/requirements.txt
|
103 |
+
2022-03-10 21:10:00,713 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
104 |
+
2022-03-10 21:10:00,714 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
105 |
+
2022-03-10 21:10:00,815 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
106 |
+
2022-03-10 21:10:00,815 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
107 |
+
2022-03-10 21:10:00,899 INFO Thread-7 :1044076 [sender.py:transition_state():392] send defer: 7
|
108 |
+
2022-03-10 21:10:00,900 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
109 |
+
2022-03-10 21:10:00,900 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 7
|
110 |
+
2022-03-10 21:10:00,900 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
111 |
+
2022-03-10 21:10:00,900 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 7
|
112 |
+
2022-03-10 21:10:00,916 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
113 |
+
2022-03-10 21:10:01,040 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 8
|
114 |
+
2022-03-10 21:10:01,040 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
115 |
+
2022-03-10 21:10:01,041 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
116 |
+
2022-03-10 21:10:01,041 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 8
|
117 |
+
2022-03-10 21:10:01,041 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
118 |
+
2022-03-10 21:10:01,041 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 8
|
119 |
+
2022-03-10 21:10:01,041 INFO SenderThread:1044076 [sender.py:transition_state():392] send defer: 9
|
120 |
+
2022-03-10 21:10:01,042 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: defer
|
121 |
+
2022-03-10 21:10:01,042 DEBUG SenderThread:1044076 [sender.py:send():235] send: final
|
122 |
+
2022-03-10 21:10:01,042 INFO HandlerThread:1044076 [handler.py:handle_request_defer():154] handle defer: 9
|
123 |
+
2022-03-10 21:10:01,043 DEBUG SenderThread:1044076 [sender.py:send():235] send: footer
|
124 |
+
2022-03-10 21:10:01,043 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: defer
|
125 |
+
2022-03-10 21:10:01,043 INFO SenderThread:1044076 [sender.py:send_request_defer():388] handle sender defer: 9
|
126 |
+
2022-03-10 21:10:01,142 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: poll_exit
|
127 |
+
2022-03-10 21:10:01,142 DEBUG SenderThread:1044076 [sender.py:send_request():249] send_request: poll_exit
|
128 |
+
2022-03-10 21:10:01,142 INFO SenderThread:1044076 [file_pusher.py:join():182] waiting for file pusher
|
129 |
+
2022-03-10 21:10:01,199 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: get_summary
|
130 |
+
2022-03-10 21:10:01,200 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: sampled_history
|
131 |
+
2022-03-10 21:10:01,201 DEBUG HandlerThread:1044076 [handler.py:handle_request():131] handle_request: shutdown
|
132 |
+
2022-03-10 21:10:01,201 INFO HandlerThread:1044076 [handler.py:finish():739] shutting down handler
|
133 |
+
2022-03-10 21:10:02,042 INFO WriterThread:1044076 [datastore.py:close():281] close: /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/run-19sg19ou.wandb
|
134 |
+
2022-03-10 21:10:02,198 INFO SenderThread:1044076 [sender.py:finish():1075] shutting down sender
|
135 |
+
2022-03-10 21:10:02,198 INFO SenderThread:1044076 [file_pusher.py:finish():177] shutting down file pusher
|
136 |
+
2022-03-10 21:10:02,198 INFO SenderThread:1044076 [file_pusher.py:join():182] waiting for file pusher
|
137 |
+
2022-03-10 21:10:02,200 INFO MainThread:1044076 [internal.py:handle_exit():79] Internal process exited
|
wandb/run-20220310_210950-19sg19ou/logs/debug.log
ADDED
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
2022-03-10 21:09:50,347 INFO MainThread:1043009 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/.config/wandb/settings
|
2 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/settings
|
3 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_setup.py:_flush():75] Loading settings from environment variables: {}
|
4 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_setup.py:_flush():75] Inferring run settings from compute environment: {'program_relpath': 'run_speech_recognition_seq2seq.py', 'program': '/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py'}
|
5 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_init.py:_log_setup():386] Logging user logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/logs/debug.log
|
6 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_init.py:_log_setup():387] Logging internal logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_210950-19sg19ou/logs/debug-internal.log
|
7 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_init.py:init():420] calling init triggers
|
8 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_init.py:init():425] wandb.init called with sweep_config: {}
|
9 |
+
config: {}
|
10 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [wandb_init.py:init():471] starting backend
|
11 |
+
2022-03-10 21:09:50,348 INFO MainThread:1043009 [backend.py:_multiprocessing_setup():99] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
|
12 |
+
2022-03-10 21:09:50,404 INFO MainThread:1043009 [backend.py:ensure_launched():219] starting backend process...
|
13 |
+
2022-03-10 21:09:50,459 INFO MainThread:1043009 [backend.py:ensure_launched():224] started backend process with pid: 1044076
|
14 |
+
2022-03-10 21:09:50,461 INFO MainThread:1043009 [wandb_init.py:init():480] backend started and connected
|
15 |
+
2022-03-10 21:09:50,471 INFO MainThread:1043009 [wandb_init.py:init():550] updated telemetry
|
16 |
+
2022-03-10 21:09:50,603 INFO MainThread:1043009 [wandb_init.py:init():581] communicating current version
|
17 |
+
2022-03-10 21:09:51,305 INFO MainThread:1043009 [wandb_init.py:init():586] got version response upgrade_message: "wandb version 0.12.11 is available! To upgrade, please run:\n $ pip install wandb --upgrade"
|
18 |
+
|
19 |
+
2022-03-10 21:09:51,305 INFO MainThread:1043009 [wandb_init.py:init():596] communicating run to backend with 30 second timeout
|
20 |
+
2022-03-10 21:09:51,398 INFO MainThread:1043009 [wandb_init.py:init():624] starting run threads in backend
|
21 |
+
2022-03-10 21:09:51,509 INFO MainThread:1043009 [wandb_run.py:_console_start():1827] atexit reg
|
22 |
+
2022-03-10 21:09:51,510 INFO MainThread:1043009 [wandb_run.py:_redirect():1701] redirect: SettingsConsole.REDIRECT
|
23 |
+
2022-03-10 21:09:51,510 INFO MainThread:1043009 [wandb_run.py:_redirect():1706] Redirecting console.
|
24 |
+
2022-03-10 21:09:51,512 INFO MainThread:1043009 [wandb_run.py:_redirect():1762] Redirects installed.
|
25 |
+
2022-03-10 21:09:51,512 INFO MainThread:1043009 [wandb_init.py:init():651] run started, returning control to user process
|
26 |
+
2022-03-10 21:09:51,515 INFO MainThread:1043009 [wandb_run.py:_config_callback():966] config_cb None None {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'torch.float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'is_encoder_decoder': True, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 50, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['SpeechEncoderDecoderModel'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 0, 'task_specific_params': None, 'problem_type': None, '_name_or_path': './', 'transformers_version': None, 'decoder': {'vocab_size': 50265, 'max_position_embeddings': 1024, 'd_model': 1024, 'encoder_ffn_dim': 4096, 'encoder_layers': 12, 'encoder_attention_heads': 16, 'decoder_ffn_dim': 4096, 'decoder_layers': 12, 'decoder_attention_heads': 16, 'dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'activation_function': 'gelu', 'init_std': 0.02, 'encoder_layerdrop': 0.0, 'decoder_layerdrop': 0.0, 'classifier_dropout': 0.0, 'use_cache': False, 'num_hidden_layers': 2, 'scale_embedding': False, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': True, 'cross_attention_hidden_size': None, 'add_cross_attention': True, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': True, 'num_beams': 4, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 3, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': 0, 'forced_eos_token_id': 2, 'remove_invalid_values': False, 'architectures': ['BartForCausalLM'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1', 2: 'LABEL_2'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1, 'LABEL_2': 2}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 0, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 2, 'task_specific_params': {'summarization': {'length_penalty': 1.0, 'max_length': 128, 'min_length': 12, 'num_beams': 4}, 'summarization_cnn': {'length_penalty': 2.0, 'max_length': 142, 'min_length': 56, 'num_beams': 4}, 'summarization_xsum': {'length_penalty': 1.0, 'max_length': 62, 'min_length': 11, 'num_beams': 6}}, 'problem_type': None, '_name_or_path': 'decoder', 'transformers_version': '4.17.0.dev0', 'add_bias_logits': False, 'add_final_layer_norm': False, 'classif_dropout': 0.0, 'gradient_checkpointing': False, 'normalize_before': False, 'model_type': 'bart'}, 'encoder': {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': None, 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['Wav2Vec2ForPreTraining'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'facebook/wav2vec2-large-lv60', 'transformers_version': '4.17.0.dev0', 'feat_extract_dropout': 0.0, 'gradient_checkpointing': False, 'hidden_dropout_prob': 0.0, 'num_feat_extract_layers': 7, 'hidden_size': 1024, 'feat_extract_norm': 'layer', 'feat_extract_activation': 'gelu', 'conv_dim': [512, 512, 512, 512, 512, 512, 512], 'conv_stride': [5, 2, 2, 2, 2, 2, 2], 'conv_kernel': [10, 3, 3, 3, 3, 2, 2], 'conv_bias': True, 'num_conv_pos_embeddings': 128, 'num_conv_pos_embedding_groups': 16, 'num_hidden_layers': 24, 'intermediate_size': 4096, 'hidden_act': 'gelu', 'num_attention_heads': 16, 'hidden_dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'feat_proj_dropout': 0.0, 'final_dropout': 0.0, 'layerdrop': 0.0, 'layer_norm_eps': 1e-05, 'initializer_range': 0.02, 'vocab_size': 32, 'do_stable_layer_norm': True, 'use_weighted_layer_sum': False, 'apply_spec_augment': False, 'mask_time_prob': 0.0, 'mask_time_length': 10, 'mask_time_min_masks': 2, 'mask_feature_prob': 0.0, 'mask_feature_length': 10, 'mask_feature_min_masks': 0, 'num_codevectors_per_group': 320, 'num_codevector_groups': 2, 'contrastive_logits_temperature': 0.1, 'feat_quantizer_dropout': 0.0, 'num_negatives': 100, 'codevector_dim': 768, 'proj_codevector_dim': 768, 'diversity_loss_weight': 0.1, 'ctc_loss_reduction': 'sum', 'ctc_zero_infinity': False, 'add_adapter': True, 'adapter_kernel_size': 3, 'adapter_stride': 2, 'num_adapter_layers': 3, 'output_hidden_size': 1024, 'classifier_proj_size': 256, 'tdnn_dim': [512, 512, 512, 512, 1500], 'tdnn_kernel': [5, 3, 3, 1, 1], 'tdnn_dilation': [1, 2, 3, 1, 1], 'xvector_output_dim': 512, 'model_type': 'wav2vec2'}, 'model_type': 'speech-encoder-decoder', 'processor_class': 'Wav2Vec2Processor', 'use_cache': False, 'output_dir': './', 'overwrite_output_dir': True, 'do_train': True, 'do_eval': True, 'do_predict': False, 'evaluation_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 16, 'per_device_eval_batch_size': 16, 'per_gpu_train_batch_size': 'None', 'per_gpu_eval_batch_size': 'None', 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': 'None', 'learning_rate': 3e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 20.0, 'max_steps': -1, 'lr_scheduler_type': 'linear', 'warmup_ratio': 0.0, 'warmup_steps': 1000, 'log_level': -1, 'log_level_replica': -1, 'log_on_each_node': True, 'logging_dir': './runs/Mar10_21-09-23_sanchit--v100', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 1500, 'save_total_limit': 1, 'save_on_each_node': False, 'no_cuda': False, 'seed': 42, 'bf16': False, 'fp16': True, 'fp16_opt_level': 'O1', 'half_precision_backend': 'amp', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': 'None', 'local_rank': -1, 'xpu_backend': 'None', 'tpu_num_cores': 'None', 'tpu_metrics_debug': False, 'debug': '[]', 'dataloader_drop_last': False, 'eval_steps': 1500, 'dataloader_num_workers': 0, 'past_index': -1, 'run_name': './', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': 'None', 'load_best_model_at_end': False, 'metric_for_best_model': 'None', 'greater_is_better': 'None', 'ignore_data_skip': False, 'sharded_ddp': '[]', 'deepspeed': 'None', 'label_smoothing_factor': 0.0, 'optim': 'adamw_hf', 'adafactor': False, 'group_by_length': True, 'length_column_name': 'input_length', 'report_to': "['tensorboard', 'wandb']", 'ddp_find_unused_parameters': 'None', 'ddp_bucket_cap_mb': 'None', 'dataloader_pin_memory': True, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': True, 'resume_from_checkpoint': 'None', 'hub_model_id': 'None', 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'gradient_checkpointing': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': 'None', 'push_to_hub_organization': 'None', 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', '_n_gpu': 1, 'mp_parameters': '', 'sortish_sampler': False, 'predict_with_generate': True, 'generation_max_length': 40, 'generation_num_beams': 1, 'train_batch_size': 16, 'eval_batch_size': 16}
|
27 |
+
2022-03-10 21:09:51,517 INFO MainThread:1043009 [wandb_watch.py:watch():43] Watching
|
28 |
+
2022-03-10 21:09:57,412 INFO MainThread:1043009 [wandb_run.py:_atexit_cleanup():1797] got exitcode: 1
|
29 |
+
2022-03-10 21:09:57,415 INFO MainThread:1043009 [wandb_run.py:_restore():1769] restore
|
30 |
+
2022-03-10 21:09:59,982 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
31 |
+
wandb_count: 1
|
32 |
+
}
|
33 |
+
pusher_stats {
|
34 |
+
uploaded_bytes: 2058
|
35 |
+
total_bytes: 2058
|
36 |
+
}
|
37 |
+
|
38 |
+
2022-03-10 21:10:00,173 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
39 |
+
wandb_count: 1
|
40 |
+
}
|
41 |
+
pusher_stats {
|
42 |
+
uploaded_bytes: 2058
|
43 |
+
total_bytes: 2058
|
44 |
+
}
|
45 |
+
|
46 |
+
2022-03-10 21:10:00,409 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
47 |
+
wandb_count: 4
|
48 |
+
}
|
49 |
+
pusher_stats {
|
50 |
+
uploaded_bytes: 2058
|
51 |
+
total_bytes: 8670
|
52 |
+
}
|
53 |
+
|
54 |
+
2022-03-10 21:10:00,511 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
55 |
+
wandb_count: 5
|
56 |
+
}
|
57 |
+
pusher_stats {
|
58 |
+
uploaded_bytes: 21382
|
59 |
+
total_bytes: 21382
|
60 |
+
}
|
61 |
+
|
62 |
+
2022-03-10 21:10:00,612 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
63 |
+
wandb_count: 5
|
64 |
+
}
|
65 |
+
pusher_stats {
|
66 |
+
uploaded_bytes: 21382
|
67 |
+
total_bytes: 21382
|
68 |
+
}
|
69 |
+
|
70 |
+
2022-03-10 21:10:00,714 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
71 |
+
wandb_count: 5
|
72 |
+
}
|
73 |
+
pusher_stats {
|
74 |
+
uploaded_bytes: 21382
|
75 |
+
total_bytes: 21382
|
76 |
+
}
|
77 |
+
|
78 |
+
2022-03-10 21:10:00,816 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
79 |
+
wandb_count: 5
|
80 |
+
}
|
81 |
+
pusher_stats {
|
82 |
+
uploaded_bytes: 21382
|
83 |
+
total_bytes: 21382
|
84 |
+
}
|
85 |
+
|
86 |
+
2022-03-10 21:10:01,041 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
|
87 |
+
wandb_count: 5
|
88 |
+
}
|
89 |
+
pusher_stats {
|
90 |
+
uploaded_bytes: 21382
|
91 |
+
total_bytes: 21382
|
92 |
+
}
|
93 |
+
|
94 |
+
2022-03-10 21:10:01,198 INFO MainThread:1043009 [wandb_run.py:_wait_for_finish():1929] got exit ret: done: true
|
95 |
+
exit_result {
|
96 |
+
}
|
97 |
+
file_counts {
|
98 |
+
wandb_count: 5
|
99 |
+
}
|
100 |
+
pusher_stats {
|
101 |
+
uploaded_bytes: 21382
|
102 |
+
total_bytes: 21382
|
103 |
+
}
|
104 |
+
local_info {
|
105 |
+
}
|
106 |
+
|
107 |
+
2022-03-10 21:10:02,333 INFO MainThread:1043009 [wandb_run.py:_append_files():2194] logging synced files
|
wandb/run-20220310_210950-19sg19ou/run-19sg19ou.wandb
ADDED
Binary file (14 kB). View file
|
wandb/run-20220310_212039-1nikmzrh/files/config.yaml
ADDED
The diff for this file is too large to render.
See raw diff
|
wandb/run-20220310_212039-1nikmzrh/files/output.log
ADDED
The diff for this file is too large to render.
See raw diff
|
wandb/run-20220310_212039-1nikmzrh/files/requirements.txt
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
absl-py==1.0.0
|
2 |
+
aiohttp==3.8.1
|
3 |
+
aiosignal==1.2.0
|
4 |
+
anyio==3.5.0
|
5 |
+
appdirs==1.4.4
|
6 |
+
argon2-cffi-bindings==21.2.0
|
7 |
+
argon2-cffi==21.3.0
|
8 |
+
asttokens==2.0.5
|
9 |
+
async-timeout==4.0.2
|
10 |
+
attrs==21.4.0
|
11 |
+
audioread==2.1.9
|
12 |
+
babel==2.9.1
|
13 |
+
backcall==0.2.0
|
14 |
+
bitsandbytes-cuda113==0.26.0
|
15 |
+
black==22.1.0
|
16 |
+
bleach==4.1.0
|
17 |
+
cachetools==5.0.0
|
18 |
+
certifi==2021.10.8
|
19 |
+
cffi==1.15.0
|
20 |
+
charset-normalizer==2.0.11
|
21 |
+
chex==0.1.0
|
22 |
+
click==8.0.3
|
23 |
+
clldutils==3.10.1
|
24 |
+
colorlog==6.6.0
|
25 |
+
csvw==1.11.0
|
26 |
+
cycler==0.11.0
|
27 |
+
datasets==1.18.3
|
28 |
+
debugpy==1.5.1
|
29 |
+
decorator==5.1.1
|
30 |
+
defusedxml==0.7.1
|
31 |
+
dill==0.3.4
|
32 |
+
dlinfo==1.2.1
|
33 |
+
dm-tree==0.1.6
|
34 |
+
docker-pycreds==0.4.0
|
35 |
+
entrypoints==0.4
|
36 |
+
executing==0.8.2
|
37 |
+
filelock==3.4.2
|
38 |
+
flatbuffers==2.0
|
39 |
+
flax==0.4.0
|
40 |
+
fonttools==4.29.1
|
41 |
+
frozenlist==1.3.0
|
42 |
+
fsspec==2022.1.0
|
43 |
+
gitdb==4.0.9
|
44 |
+
gitpython==3.1.27
|
45 |
+
google-auth-oauthlib==0.4.6
|
46 |
+
google-auth==2.6.0
|
47 |
+
grpcio==1.43.0
|
48 |
+
huggingface-hub==0.4.0
|
49 |
+
hypothesis==6.36.1
|
50 |
+
idna==3.3
|
51 |
+
importlib-metadata==4.10.1
|
52 |
+
ipykernel==6.8.0
|
53 |
+
ipython-genutils==0.2.0
|
54 |
+
ipython==8.0.1
|
55 |
+
ipywidgets==7.6.5
|
56 |
+
isodate==0.6.1
|
57 |
+
jax==0.2.28
|
58 |
+
jaxlib==0.1.76+cuda11.cudnn82
|
59 |
+
jedi==0.18.1
|
60 |
+
jinja2==3.0.3
|
61 |
+
jiwer==2.3.0
|
62 |
+
joblib==1.1.0
|
63 |
+
json5==0.9.6
|
64 |
+
jsonschema==4.4.0
|
65 |
+
jupyter-client==7.1.2
|
66 |
+
jupyter-console==6.4.0
|
67 |
+
jupyter-core==4.9.1
|
68 |
+
jupyter-server==1.13.5
|
69 |
+
jupyter==1.0.0
|
70 |
+
jupyterlab-pygments==0.1.2
|
71 |
+
jupyterlab-server==2.10.3
|
72 |
+
jupyterlab-widgets==1.0.2
|
73 |
+
jupyterlab==3.2.9
|
74 |
+
kiwisolver==1.3.2
|
75 |
+
librosa==0.8.1
|
76 |
+
llvmlite==0.38.0
|
77 |
+
markdown==3.3.6
|
78 |
+
markupsafe==2.0.1
|
79 |
+
matplotlib-inline==0.1.3
|
80 |
+
matplotlib==3.5.1
|
81 |
+
mistune==0.8.4
|
82 |
+
msgpack==1.0.3
|
83 |
+
multidict==6.0.2
|
84 |
+
multiprocess==0.70.12.2
|
85 |
+
mypy-extensions==0.4.3
|
86 |
+
nbclassic==0.3.5
|
87 |
+
nbclient==0.5.10
|
88 |
+
nbconvert==6.4.1
|
89 |
+
nbformat==5.1.3
|
90 |
+
nest-asyncio==1.5.4
|
91 |
+
notebook==6.4.8
|
92 |
+
numba==0.55.1
|
93 |
+
numpy==1.21.5
|
94 |
+
oauthlib==3.2.0
|
95 |
+
opt-einsum==3.3.0
|
96 |
+
optax==0.1.0
|
97 |
+
packaging==21.3
|
98 |
+
pandas==1.4.0
|
99 |
+
pandocfilters==1.5.0
|
100 |
+
parso==0.8.3
|
101 |
+
pathspec==0.9.0
|
102 |
+
pathtools==0.1.2
|
103 |
+
pexpect==4.8.0
|
104 |
+
phonemizer==3.0.1
|
105 |
+
pickleshare==0.7.5
|
106 |
+
pillow==9.0.0
|
107 |
+
pip==22.0.2
|
108 |
+
pkg-resources==0.0.0
|
109 |
+
platformdirs==2.4.1
|
110 |
+
pooch==1.6.0
|
111 |
+
prometheus-client==0.13.1
|
112 |
+
promise==2.3
|
113 |
+
prompt-toolkit==3.0.26
|
114 |
+
protobuf==3.19.4
|
115 |
+
psutil==5.9.0
|
116 |
+
ptyprocess==0.7.0
|
117 |
+
pure-eval==0.2.2
|
118 |
+
pyarrow==6.0.1
|
119 |
+
pyasn1-modules==0.2.8
|
120 |
+
pyasn1==0.4.8
|
121 |
+
pycparser==2.21
|
122 |
+
pyctcdecode==0.3.0
|
123 |
+
pygments==2.11.2
|
124 |
+
pygtrie==2.4.2
|
125 |
+
pyparsing==3.0.7
|
126 |
+
pyrsistent==0.18.1
|
127 |
+
python-dateutil==2.8.2
|
128 |
+
python-levenshtein==0.12.2
|
129 |
+
pytz==2021.3
|
130 |
+
pyyaml==6.0
|
131 |
+
pyzmq==22.3.0
|
132 |
+
qtconsole==5.2.2
|
133 |
+
qtpy==2.0.1
|
134 |
+
regex==2022.1.18
|
135 |
+
requests-oauthlib==1.3.1
|
136 |
+
requests==2.27.1
|
137 |
+
resampy==0.2.2
|
138 |
+
rfc3986==2.0.0
|
139 |
+
rsa==4.8
|
140 |
+
sacremoses==0.0.47
|
141 |
+
scikit-learn==1.0.2
|
142 |
+
scipy==1.7.3
|
143 |
+
segments==2.2.0
|
144 |
+
send2trash==1.8.0
|
145 |
+
sentry-sdk==1.5.6
|
146 |
+
setuptools==44.1.1
|
147 |
+
shortuuid==1.0.8
|
148 |
+
six==1.16.0
|
149 |
+
smmap==5.0.0
|
150 |
+
sniffio==1.2.0
|
151 |
+
sortedcontainers==2.4.0
|
152 |
+
soundfile==0.10.3.post1
|
153 |
+
stack-data==0.1.4
|
154 |
+
tabulate==0.8.9
|
155 |
+
tensorboard-data-server==0.6.1
|
156 |
+
tensorboard-plugin-wit==1.8.1
|
157 |
+
tensorboard==2.8.0
|
158 |
+
termcolor==1.1.0
|
159 |
+
terminado==0.13.1
|
160 |
+
testpath==0.5.0
|
161 |
+
threadpoolctl==3.1.0
|
162 |
+
tokenizers==0.11.4
|
163 |
+
tomli==2.0.0
|
164 |
+
toolz==0.11.2
|
165 |
+
torch==1.10.2+cu113
|
166 |
+
torchaudio==0.10.2+cu113
|
167 |
+
tornado==6.1
|
168 |
+
tqdm==4.62.3
|
169 |
+
traitlets==5.1.1
|
170 |
+
transformers==4.17.0.dev0
|
171 |
+
typing-extensions==3.10.0.2
|
172 |
+
uritemplate==4.1.1
|
173 |
+
urllib3==1.26.8
|
174 |
+
wandb==0.12.10
|
175 |
+
wcwidth==0.2.5
|
176 |
+
webencodings==0.5.1
|
177 |
+
websocket-client==1.2.3
|
178 |
+
werkzeug==2.0.2
|
179 |
+
wheel==0.37.1
|
180 |
+
widgetsnbextension==3.5.2
|
181 |
+
xxhash==2.0.2
|
182 |
+
yarl==1.7.2
|
183 |
+
yaspin==2.1.0
|
184 |
+
zipp==3.7.0
|
wandb/run-20220310_212039-1nikmzrh/files/wandb-metadata.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"os": "Linux-5.11.0-1028-gcp-x86_64-with-glibc2.33",
|
3 |
+
"python": "3.9.5",
|
4 |
+
"heartbeatAt": "2022-03-10T21:20:40.271786",
|
5 |
+
"startedAt": "2022-03-10T21:20:39.161849",
|
6 |
+
"docker": null,
|
7 |
+
"gpu": "Tesla V100-SXM2-16GB",
|
8 |
+
"gpu_count": 2,
|
9 |
+
"cpu_count": 16,
|
10 |
+
"cuda": null,
|
11 |
+
"args": [
|
12 |
+
"--dataset_name=librispeech_asr",
|
13 |
+
"--model_name_or_path=./",
|
14 |
+
"--tokenizer_name=./",
|
15 |
+
"--dataset_config_name=clean",
|
16 |
+
"--train_split_name=train.100",
|
17 |
+
"--eval_split_name=validation",
|
18 |
+
"--output_dir=./",
|
19 |
+
"--preprocessing_num_workers=1",
|
20 |
+
"--length_column_name=input_length",
|
21 |
+
"--overwrite_output_dir",
|
22 |
+
"--num_train_epochs=20",
|
23 |
+
"--per_device_train_batch_size=8",
|
24 |
+
"--per_device_eval_batch_size=8",
|
25 |
+
"--gradient_accumulation_steps=4",
|
26 |
+
"--generation_max_length=40",
|
27 |
+
"--generation_num_beams=1",
|
28 |
+
"--learning_rate=3e-5",
|
29 |
+
"--warmup_steps=1000",
|
30 |
+
"--evaluation_strategy=steps",
|
31 |
+
"--text_column_name=text",
|
32 |
+
"--save_steps=1500",
|
33 |
+
"--eval_steps=1500",
|
34 |
+
"--logging_steps=1",
|
35 |
+
"--save_total_limit=1",
|
36 |
+
"--freeze_feature_encoder",
|
37 |
+
"--gradient_checkpointing",
|
38 |
+
"--fp16",
|
39 |
+
"--group_by_length",
|
40 |
+
"--predict_with_generate",
|
41 |
+
"--do_lower_case",
|
42 |
+
"--do_eval",
|
43 |
+
"--do_train",
|
44 |
+
"--push_to_hub",
|
45 |
+
"--use_auth_token"
|
46 |
+
],
|
47 |
+
"state": "running",
|
48 |
+
"program": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py",
|
49 |
+
"codePath": "run_speech_recognition_seq2seq.py",
|
50 |
+
"git": {
|
51 |
+
"remote": "https://huggingface.co/sanchit-gandhi/wav2vec2-2-rnd-2-layer-bart",
|
52 |
+
"commit": "4ea7f446103e7b0489f13a4de255ec90e6838d3b"
|
53 |
+
},
|
54 |
+
"email": "sanchit@huggingface.co",
|
55 |
+
"root": "/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart",
|
56 |
+
"host": "sanchit--v100",
|
57 |
+
"username": "sanchit_huggingface_co",
|
58 |
+
"executable": "/home/sanchit_huggingface_co/gcp/bin/python"
|
59 |
+
}
|
wandb/run-20220310_212039-1nikmzrh/files/wandb-summary.json
ADDED
The diff for this file is too large to render.
See raw diff
|
wandb/run-20220310_212039-1nikmzrh/logs/debug-internal.log
ADDED
The diff for this file is too large to render.
See raw diff
|
wandb/run-20220310_212039-1nikmzrh/logs/debug.log
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
2022-03-10 21:20:39,163 INFO MainThread:1065811 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/.config/wandb/settings
|
2 |
+
2022-03-10 21:20:39,163 INFO MainThread:1065811 [wandb_setup.py:_flush():75] Loading settings from /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/settings
|
3 |
+
2022-03-10 21:20:39,163 INFO MainThread:1065811 [wandb_setup.py:_flush():75] Loading settings from environment variables: {}
|
4 |
+
2022-03-10 21:20:39,163 INFO MainThread:1065811 [wandb_setup.py:_flush():75] Inferring run settings from compute environment: {'program_relpath': 'run_speech_recognition_seq2seq.py', 'program': '/home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/run_speech_recognition_seq2seq.py'}
|
5 |
+
2022-03-10 21:20:39,163 INFO MainThread:1065811 [wandb_init.py:_log_setup():386] Logging user logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_212039-1nikmzrh/logs/debug.log
|
6 |
+
2022-03-10 21:20:39,164 INFO MainThread:1065811 [wandb_init.py:_log_setup():387] Logging internal logs to /home/sanchit_huggingface_co/wav2vec2-2-rnd-2-layer-bart/wandb/run-20220310_212039-1nikmzrh/logs/debug-internal.log
|
7 |
+
2022-03-10 21:20:39,164 INFO MainThread:1065811 [wandb_init.py:init():420] calling init triggers
|
8 |
+
2022-03-10 21:20:39,164 INFO MainThread:1065811 [wandb_init.py:init():425] wandb.init called with sweep_config: {}
|
9 |
+
config: {}
|
10 |
+
2022-03-10 21:20:39,164 INFO MainThread:1065811 [wandb_init.py:init():471] starting backend
|
11 |
+
2022-03-10 21:20:39,164 INFO MainThread:1065811 [backend.py:_multiprocessing_setup():99] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
|
12 |
+
2022-03-10 21:20:39,222 INFO MainThread:1065811 [backend.py:ensure_launched():219] starting backend process...
|
13 |
+
2022-03-10 21:20:39,278 INFO MainThread:1065811 [backend.py:ensure_launched():224] started backend process with pid: 1066879
|
14 |
+
2022-03-10 21:20:39,280 INFO MainThread:1065811 [wandb_init.py:init():480] backend started and connected
|
15 |
+
2022-03-10 21:20:39,290 INFO MainThread:1065811 [wandb_init.py:init():550] updated telemetry
|
16 |
+
2022-03-10 21:20:39,429 INFO MainThread:1065811 [wandb_init.py:init():581] communicating current version
|
17 |
+
2022-03-10 21:20:40,179 INFO MainThread:1065811 [wandb_init.py:init():586] got version response upgrade_message: "wandb version 0.12.11 is available! To upgrade, please run:\n $ pip install wandb --upgrade"
|
18 |
+
|
19 |
+
2022-03-10 21:20:40,179 INFO MainThread:1065811 [wandb_init.py:init():596] communicating run to backend with 30 second timeout
|
20 |
+
2022-03-10 21:20:40,266 INFO MainThread:1065811 [wandb_init.py:init():624] starting run threads in backend
|
21 |
+
2022-03-10 21:20:40,380 INFO MainThread:1065811 [wandb_run.py:_console_start():1827] atexit reg
|
22 |
+
2022-03-10 21:20:40,381 INFO MainThread:1065811 [wandb_run.py:_redirect():1701] redirect: SettingsConsole.REDIRECT
|
23 |
+
2022-03-10 21:20:40,381 INFO MainThread:1065811 [wandb_run.py:_redirect():1706] Redirecting console.
|
24 |
+
2022-03-10 21:20:40,383 INFO MainThread:1065811 [wandb_run.py:_redirect():1762] Redirects installed.
|
25 |
+
2022-03-10 21:20:40,383 INFO MainThread:1065811 [wandb_init.py:init():651] run started, returning control to user process
|
26 |
+
2022-03-10 21:20:40,387 INFO MainThread:1065811 [wandb_run.py:_config_callback():966] config_cb None None {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'torch.float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'is_encoder_decoder': True, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 50, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['SpeechEncoderDecoderModel'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 0, 'task_specific_params': None, 'problem_type': None, '_name_or_path': './', 'transformers_version': None, 'decoder': {'vocab_size': 50265, 'max_position_embeddings': 1024, 'd_model': 1024, 'encoder_ffn_dim': 4096, 'encoder_layers': 12, 'encoder_attention_heads': 16, 'decoder_ffn_dim': 4096, 'decoder_layers': 12, 'decoder_attention_heads': 16, 'dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'activation_function': 'gelu', 'init_std': 0.02, 'encoder_layerdrop': 0.0, 'decoder_layerdrop': 0.0, 'classifier_dropout': 0.0, 'use_cache': False, 'num_hidden_layers': 2, 'scale_embedding': False, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': True, 'cross_attention_hidden_size': None, 'add_cross_attention': True, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': True, 'num_beams': 4, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 3, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': 0, 'forced_eos_token_id': 2, 'remove_invalid_values': False, 'architectures': ['BartForCausalLM'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1', 2: 'LABEL_2'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1, 'LABEL_2': 2}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 0, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 2, 'task_specific_params': {'summarization': {'length_penalty': 1.0, 'max_length': 128, 'min_length': 12, 'num_beams': 4}, 'summarization_cnn': {'length_penalty': 2.0, 'max_length': 142, 'min_length': 56, 'num_beams': 4}, 'summarization_xsum': {'length_penalty': 1.0, 'max_length': 62, 'min_length': 11, 'num_beams': 6}}, 'problem_type': None, '_name_or_path': 'decoder', 'transformers_version': '4.17.0.dev0', 'add_bias_logits': False, 'add_final_layer_norm': False, 'classif_dropout': 0.0, 'gradient_checkpointing': False, 'normalize_before': False, 'model_type': 'bart'}, 'encoder': {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': None, 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['Wav2Vec2ForPreTraining'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'facebook/wav2vec2-large-lv60', 'transformers_version': '4.17.0.dev0', 'feat_extract_dropout': 0.0, 'gradient_checkpointing': False, 'hidden_dropout_prob': 0.0, 'num_feat_extract_layers': 7, 'hidden_size': 1024, 'feat_extract_norm': 'layer', 'feat_extract_activation': 'gelu', 'conv_dim': [512, 512, 512, 512, 512, 512, 512], 'conv_stride': [5, 2, 2, 2, 2, 2, 2], 'conv_kernel': [10, 3, 3, 3, 3, 2, 2], 'conv_bias': True, 'num_conv_pos_embeddings': 128, 'num_conv_pos_embedding_groups': 16, 'num_hidden_layers': 24, 'intermediate_size': 4096, 'hidden_act': 'gelu', 'num_attention_heads': 16, 'hidden_dropout': 0.0, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'feat_proj_dropout': 0.0, 'final_dropout': 0.0, 'layerdrop': 0.0, 'layer_norm_eps': 1e-05, 'initializer_range': 0.02, 'vocab_size': 32, 'do_stable_layer_norm': True, 'use_weighted_layer_sum': False, 'apply_spec_augment': False, 'mask_time_prob': 0.0, 'mask_time_length': 10, 'mask_time_min_masks': 2, 'mask_feature_prob': 0.0, 'mask_feature_length': 10, 'mask_feature_min_masks': 0, 'num_codevectors_per_group': 320, 'num_codevector_groups': 2, 'contrastive_logits_temperature': 0.1, 'feat_quantizer_dropout': 0.0, 'num_negatives': 100, 'codevector_dim': 768, 'proj_codevector_dim': 768, 'diversity_loss_weight': 0.1, 'ctc_loss_reduction': 'sum', 'ctc_zero_infinity': False, 'add_adapter': True, 'adapter_kernel_size': 3, 'adapter_stride': 2, 'num_adapter_layers': 3, 'output_hidden_size': 1024, 'classifier_proj_size': 256, 'tdnn_dim': [512, 512, 512, 512, 1500], 'tdnn_kernel': [5, 3, 3, 1, 1], 'tdnn_dilation': [1, 2, 3, 1, 1], 'xvector_output_dim': 512, 'model_type': 'wav2vec2'}, 'model_type': 'speech-encoder-decoder', 'processor_class': 'Wav2Vec2Processor', 'use_cache': False, 'output_dir': './', 'overwrite_output_dir': True, 'do_train': True, 'do_eval': True, 'do_predict': False, 'evaluation_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 8, 'per_device_eval_batch_size': 8, 'per_gpu_train_batch_size': 'None', 'per_gpu_eval_batch_size': 'None', 'gradient_accumulation_steps': 4, 'eval_accumulation_steps': 'None', 'learning_rate': 3e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 20.0, 'max_steps': -1, 'lr_scheduler_type': 'linear', 'warmup_ratio': 0.0, 'warmup_steps': 1000, 'log_level': -1, 'log_level_replica': -1, 'log_on_each_node': True, 'logging_dir': './runs/Mar10_21-20-12_sanchit--v100', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 1500, 'save_total_limit': 1, 'save_on_each_node': False, 'no_cuda': False, 'seed': 42, 'bf16': False, 'fp16': True, 'fp16_opt_level': 'O1', 'half_precision_backend': 'amp', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': 'None', 'local_rank': -1, 'xpu_backend': 'None', 'tpu_num_cores': 'None', 'tpu_metrics_debug': False, 'debug': '[]', 'dataloader_drop_last': False, 'eval_steps': 1500, 'dataloader_num_workers': 0, 'past_index': -1, 'run_name': './', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': 'None', 'load_best_model_at_end': False, 'metric_for_best_model': 'None', 'greater_is_better': 'None', 'ignore_data_skip': False, 'sharded_ddp': '[]', 'deepspeed': 'None', 'label_smoothing_factor': 0.0, 'optim': 'adamw_hf', 'adafactor': False, 'group_by_length': True, 'length_column_name': 'input_length', 'report_to': "['tensorboard', 'wandb']", 'ddp_find_unused_parameters': 'None', 'ddp_bucket_cap_mb': 'None', 'dataloader_pin_memory': True, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': True, 'resume_from_checkpoint': 'None', 'hub_model_id': 'None', 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'gradient_checkpointing': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': 'None', 'push_to_hub_organization': 'None', 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', '_n_gpu': 1, 'mp_parameters': '', 'sortish_sampler': False, 'predict_with_generate': True, 'generation_max_length': 40, 'generation_num_beams': 1, 'train_batch_size': 8, 'eval_batch_size': 8}
|
27 |
+
2022-03-10 21:20:40,392 INFO MainThread:1065811 [wandb_watch.py:watch():43] Watching
|
wandb/run-20220310_212039-1nikmzrh/run-1nikmzrh.wandb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:da556577d956430417dbc232c155e2e945407b22dffaa37ab8d740bbf1565f60
|
3 |
+
size 75371130
|