noflm
commited on
Commit
•
d922f90
1
Parent(s):
86574ed
Training in progress, step 8000
Browse files- checkpoint-8000/config.json +54 -0
- checkpoint-8000/optimizer.pt +3 -0
- checkpoint-8000/preprocessor_config.json +0 -0
- checkpoint-8000/pytorch_model.bin +3 -0
- checkpoint-8000/rng_state.pth +3 -0
- checkpoint-8000/scaler.pt +3 -0
- checkpoint-8000/scheduler.pt +3 -0
- checkpoint-8000/trainer_state.json +184 -0
- checkpoint-8000/training_args.bin +3 -0
- pytorch_model.bin +1 -1
- runs/Dec27_09-55-14_srv/events.out.tfevents.1672134971.srv.1172111.0 +2 -2
checkpoint-8000/config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "openai/whisper-base",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "gelu",
|
5 |
+
"architectures": [
|
6 |
+
"WhisperForConditionalGeneration"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.0,
|
9 |
+
"begin_suppress_tokens": [
|
10 |
+
220,
|
11 |
+
50257
|
12 |
+
],
|
13 |
+
"bos_token_id": 50257,
|
14 |
+
"d_model": 512,
|
15 |
+
"decoder_attention_heads": 8,
|
16 |
+
"decoder_ffn_dim": 2048,
|
17 |
+
"decoder_layerdrop": 0.0,
|
18 |
+
"decoder_layers": 6,
|
19 |
+
"decoder_start_token_id": 50258,
|
20 |
+
"dropout": 0.0,
|
21 |
+
"encoder_attention_heads": 8,
|
22 |
+
"encoder_ffn_dim": 2048,
|
23 |
+
"encoder_layerdrop": 0.0,
|
24 |
+
"encoder_layers": 6,
|
25 |
+
"eos_token_id": 50257,
|
26 |
+
"forced_decoder_ids": [
|
27 |
+
[
|
28 |
+
1,
|
29 |
+
50266
|
30 |
+
],
|
31 |
+
[
|
32 |
+
2,
|
33 |
+
50359
|
34 |
+
],
|
35 |
+
[
|
36 |
+
3,
|
37 |
+
50363
|
38 |
+
]
|
39 |
+
],
|
40 |
+
"init_std": 0.02,
|
41 |
+
"is_encoder_decoder": true,
|
42 |
+
"max_length": 448,
|
43 |
+
"max_source_positions": 1500,
|
44 |
+
"max_target_positions": 448,
|
45 |
+
"model_type": "whisper",
|
46 |
+
"num_hidden_layers": 6,
|
47 |
+
"num_mel_bins": 80,
|
48 |
+
"pad_token_id": 50257,
|
49 |
+
"scale_embedding": false,
|
50 |
+
"torch_dtype": "float32",
|
51 |
+
"transformers_version": "4.26.0.dev0",
|
52 |
+
"use_cache": false,
|
53 |
+
"vocab_size": 51865
|
54 |
+
}
|
checkpoint-8000/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:77ada3dc2330e192fa51c9c14c5059a59d4b64b65b9af2da232c5f88f0c3411f
|
3 |
+
size 580895237
|
checkpoint-8000/preprocessor_config.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-8000/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b39d30a761cef641b63b89e6f34579521f1660e00d2a05680556d92eaed894eb
|
3 |
+
size 290458721
|
checkpoint-8000/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:33dcf37de5495c2285c9677c81c6207859032779e9c4ed30e619d170591b1ba7
|
3 |
+
size 14639
|
checkpoint-8000/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e8d407729cf398de31e94e18fff7f357437646319b163e49b1d38c97af5554b
|
3 |
+
size 557
|
checkpoint-8000/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7d67fa48d23d025c52e6000e3850175b41e26849294e0760187728015af5bc54
|
3 |
+
size 627
|
checkpoint-8000/trainer_state.json
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 28.0132,
|
5 |
+
"global_step": 8000,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 1.02,
|
12 |
+
"learning_rate": 9.960000000000001e-06,
|
13 |
+
"loss": 0.5615,
|
14 |
+
"step": 500
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 3.02,
|
18 |
+
"learning_rate": 1e-05,
|
19 |
+
"loss": 0.3273,
|
20 |
+
"step": 1000
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 3.02,
|
24 |
+
"eval_loss": 0.4225325584411621,
|
25 |
+
"eval_runtime": 1489.9228,
|
26 |
+
"eval_samples_per_second": 3.09,
|
27 |
+
"eval_steps_per_second": 0.193,
|
28 |
+
"eval_wer": 20.82533081951615,
|
29 |
+
"step": 1000
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"epoch": 5.01,
|
33 |
+
"learning_rate": 1e-05,
|
34 |
+
"loss": 0.1859,
|
35 |
+
"step": 1500
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"epoch": 7.0,
|
39 |
+
"learning_rate": 1e-05,
|
40 |
+
"loss": 0.0923,
|
41 |
+
"step": 2000
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 7.0,
|
45 |
+
"eval_loss": 0.46434369683265686,
|
46 |
+
"eval_runtime": 1504.5618,
|
47 |
+
"eval_samples_per_second": 3.06,
|
48 |
+
"eval_steps_per_second": 0.191,
|
49 |
+
"eval_wer": 21.22002361108664,
|
50 |
+
"step": 2000
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 8.03,
|
54 |
+
"learning_rate": 1e-05,
|
55 |
+
"loss": 0.0369,
|
56 |
+
"step": 2500
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 10.02,
|
60 |
+
"learning_rate": 1e-05,
|
61 |
+
"loss": 0.0164,
|
62 |
+
"step": 3000
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"epoch": 10.02,
|
66 |
+
"eval_loss": 0.5403426885604858,
|
67 |
+
"eval_runtime": 1504.7378,
|
68 |
+
"eval_samples_per_second": 3.06,
|
69 |
+
"eval_steps_per_second": 0.191,
|
70 |
+
"eval_wer": 22.962662766725987,
|
71 |
+
"step": 3000
|
72 |
+
},
|
73 |
+
{
|
74 |
+
"epoch": 12.01,
|
75 |
+
"learning_rate": 1e-05,
|
76 |
+
"loss": 0.0089,
|
77 |
+
"step": 3500
|
78 |
+
},
|
79 |
+
{
|
80 |
+
"epoch": 14.01,
|
81 |
+
"learning_rate": 1e-05,
|
82 |
+
"loss": 0.006,
|
83 |
+
"step": 4000
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 14.01,
|
87 |
+
"eval_loss": 0.5819889307022095,
|
88 |
+
"eval_runtime": 1478.853,
|
89 |
+
"eval_samples_per_second": 3.113,
|
90 |
+
"eval_steps_per_second": 0.195,
|
91 |
+
"eval_wer": 21.086109985375224,
|
92 |
+
"step": 4000
|
93 |
+
},
|
94 |
+
{
|
95 |
+
"epoch": 16.0,
|
96 |
+
"learning_rate": 1e-05,
|
97 |
+
"loss": 0.0057,
|
98 |
+
"step": 4500
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 17.02,
|
102 |
+
"learning_rate": 1e-05,
|
103 |
+
"loss": 0.0046,
|
104 |
+
"step": 5000
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 17.02,
|
108 |
+
"eval_loss": 0.5851565003395081,
|
109 |
+
"eval_runtime": 1532.3442,
|
110 |
+
"eval_samples_per_second": 3.005,
|
111 |
+
"eval_steps_per_second": 0.188,
|
112 |
+
"eval_wer": 22.072841964301446,
|
113 |
+
"step": 5000
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 19.02,
|
117 |
+
"learning_rate": 1e-05,
|
118 |
+
"loss": 0.0042,
|
119 |
+
"step": 5500
|
120 |
+
},
|
121 |
+
{
|
122 |
+
"epoch": 21.01,
|
123 |
+
"learning_rate": 1e-05,
|
124 |
+
"loss": 0.0034,
|
125 |
+
"step": 6000
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"epoch": 21.01,
|
129 |
+
"eval_loss": 0.6112661361694336,
|
130 |
+
"eval_runtime": 1497.7325,
|
131 |
+
"eval_samples_per_second": 3.074,
|
132 |
+
"eval_steps_per_second": 0.192,
|
133 |
+
"eval_wer": 21.6622909802125,
|
134 |
+
"step": 6000
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 23.0,
|
138 |
+
"learning_rate": 1e-05,
|
139 |
+
"loss": 0.004,
|
140 |
+
"step": 6500
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 24.03,
|
144 |
+
"learning_rate": 1e-05,
|
145 |
+
"loss": 0.0028,
|
146 |
+
"step": 7000
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 24.03,
|
150 |
+
"eval_loss": 0.6582160592079163,
|
151 |
+
"eval_runtime": 1517.5965,
|
152 |
+
"eval_samples_per_second": 3.034,
|
153 |
+
"eval_steps_per_second": 0.19,
|
154 |
+
"eval_wer": 22.326573044596763,
|
155 |
+
"step": 7000
|
156 |
+
},
|
157 |
+
{
|
158 |
+
"epoch": 26.02,
|
159 |
+
"learning_rate": 1e-05,
|
160 |
+
"loss": 0.0027,
|
161 |
+
"step": 7500
|
162 |
+
},
|
163 |
+
{
|
164 |
+
"epoch": 28.01,
|
165 |
+
"learning_rate": 1e-05,
|
166 |
+
"loss": 0.0025,
|
167 |
+
"step": 8000
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 28.01,
|
171 |
+
"eval_loss": 0.6349595785140991,
|
172 |
+
"eval_runtime": 1539.4689,
|
173 |
+
"eval_samples_per_second": 2.991,
|
174 |
+
"eval_steps_per_second": 0.187,
|
175 |
+
"eval_wer": 22.23318591087696,
|
176 |
+
"step": 8000
|
177 |
+
}
|
178 |
+
],
|
179 |
+
"max_steps": 10000,
|
180 |
+
"num_train_epochs": 9223372036854775807,
|
181 |
+
"total_flos": 1.656420525932544e+19,
|
182 |
+
"trial_name": null,
|
183 |
+
"trial_params": null
|
184 |
+
}
|
checkpoint-8000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:63250d2e981e0cfad1a5508019d2b4e5ef4e64d9d02e786f5ee65312228c5eb2
|
3 |
+
size 3643
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 290458721
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b39d30a761cef641b63b89e6f34579521f1660e00d2a05680556d92eaed894eb
|
3 |
size 290458721
|
runs/Dec27_09-55-14_srv/events.out.tfevents.1672134971.srv.1172111.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7572d937da3a10588e85043a93ac82645e24986b1c03d00661facfeea4c4027f
|
3 |
+
size 9484
|