nityanandmathur commited on
Commit
b267c70
·
verified ·
1 Parent(s): e146f4b

add B5_2 (completed, step 30000)

Browse files
Files changed (3) hide show
  1. B5_2/config.json +4 -4
  2. B5_2/model.safetensors +1 -1
  3. B5_2/run.json +163 -2
B5_2/config.json CHANGED
@@ -9,11 +9,11 @@
9
  "total_params": 65757696,
10
  "mup_base_width": 256,
11
  "lr_base": 0.004,
12
- "steps": 5000,
13
- "status": "running",
14
  "codec": "kyutai/mimi (12.5 Hz, 8 codebooks, 2048+MASK+PAD)",
15
  "phoneme_vocab_file": "phone_vocab.json",
16
- "final_val_loss": null,
17
- "gpu_hours": null,
18
  "dtype": "bfloat16"
19
  }
 
9
  "total_params": 65757696,
10
  "mup_base_width": 256,
11
  "lr_base": 0.004,
12
+ "steps": 30000,
13
+ "status": "completed",
14
  "codec": "kyutai/mimi (12.5 Hz, 8 codebooks, 2048+MASK+PAD)",
15
  "phoneme_vocab_file": "phone_vocab.json",
16
+ "final_val_loss": 4.529994341288424,
17
+ "gpu_hours": 4.284989529450734,
18
  "dtype": "bfloat16"
19
  }
B5_2/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ca490db7c22bc69df8911019c1ce3880106e93f3e2e7bfc3cc0b7c4a5248adf1
3
  size 131533696
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee922bfe40172f847bc1e559c2f20820ba396e395a8fca91790df9947e5b0a2e
3
  size 131533696
B5_2/run.json CHANGED
@@ -11,6 +11,167 @@
11
  "micro_batch": 64,
12
  "accum": 4,
13
  "device": "NVIDIA B200",
14
- "started": 1786011576.1702216,
15
- "status": "running"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
16
  }
 
11
  "micro_batch": 64,
12
  "accum": 4,
13
  "device": "NVIDIA B200",
14
+ "started": 1786018601.3426707,
15
+ "status": "completed",
16
+ "recipe": "coarse",
17
+ "wall_seconds": 15425.962306022644,
18
+ "gpu_hours": 4.284989529450734,
19
+ "wall_seconds_this_segment": 6379.65832567215,
20
+ "final_step": 30000,
21
+ "val_hist": [
22
+ {
23
+ "step": 1000,
24
+ "val_loss": 6.14543048990591,
25
+ "wall": 458.56310963630676
26
+ },
27
+ {
28
+ "step": 2000,
29
+ "val_loss": 5.8506344698386235,
30
+ "wall": 805.4449908733368
31
+ },
32
+ {
33
+ "step": 3000,
34
+ "val_loss": 5.675811799313059,
35
+ "wall": 1369.2264199256897
36
+ },
37
+ {
38
+ "step": 4000,
39
+ "val_loss": 5.559969626347653,
40
+ "wall": 2210.125162601471
41
+ },
42
+ {
43
+ "step": 5000,
44
+ "val_loss": 5.449884404870357,
45
+ "wall": 2952.8852343559265
46
+ },
47
+ {
48
+ "step": 6000,
49
+ "val_loss": 5.371254819607582,
50
+ "wall": 3753.7409160137177
51
+ },
52
+ {
53
+ "step": 7000,
54
+ "val_loss": 5.306319362421569,
55
+ "wall": 4508.262000799179
56
+ },
57
+ {
58
+ "step": 8000,
59
+ "val_loss": 5.244359692632699,
60
+ "wall": 5278.430006504059
61
+ },
62
+ {
63
+ "step": 9000,
64
+ "val_loss": 5.191215040974669,
65
+ "wall": 6029.729672193527
66
+ },
67
+ {
68
+ "step": 10000,
69
+ "val_loss": 5.129874188900989,
70
+ "wall": 6798.5870134830475
71
+ },
72
+ {
73
+ "step": 11000,
74
+ "val_loss": 5.087892140533566,
75
+ "wall": 7546.976212501526
76
+ },
77
+ {
78
+ "step": 12000,
79
+ "val_loss": 5.046070887618726,
80
+ "wall": 8297.954487085342
81
+ },
82
+ {
83
+ "step": 13000,
84
+ "val_loss": 5.012589050460491,
85
+ "wall": 9046.302416324615
86
+ },
87
+ {
88
+ "step": 14000,
89
+ "val_loss": 4.982054855000495,
90
+ "wall": 376.37985253334045
91
+ },
92
+ {
93
+ "step": 15000,
94
+ "val_loss": 4.949626044818139,
95
+ "wall": 730.476071357727
96
+ },
97
+ {
98
+ "step": 16000,
99
+ "val_loss": 4.921173386013243,
100
+ "wall": 1081.6135883331299
101
+ },
102
+ {
103
+ "step": 17000,
104
+ "val_loss": 4.89383888840242,
105
+ "wall": 1542.9732348918915
106
+ },
107
+ {
108
+ "step": 18000,
109
+ "val_loss": 4.854102222785649,
110
+ "wall": 1889.135976076126
111
+ },
112
+ {
113
+ "step": 19000,
114
+ "val_loss": 4.824655856002446,
115
+ "wall": 2238.154031276703
116
+ },
117
+ {
118
+ "step": 20000,
119
+ "val_loss": 4.7921797886163,
120
+ "wall": 2584.7983260154724
121
+ },
122
+ {
123
+ "step": 21000,
124
+ "val_loss": 4.757751051963086,
125
+ "wall": 2931.764621257782
126
+ },
127
+ {
128
+ "step": 22000,
129
+ "val_loss": 4.724672726666488,
130
+ "wall": 3319.084920167923
131
+ },
132
+ {
133
+ "step": 23000,
134
+ "val_loss": 4.691025784324143,
135
+ "wall": 3705.8441939353943
136
+ },
137
+ {
138
+ "step": 24000,
139
+ "val_loss": 4.6580051455340925,
140
+ "wall": 4086.6612861156464
141
+ },
142
+ {
143
+ "step": 25000,
144
+ "val_loss": 4.625841556524655,
145
+ "wall": 4539.412252187729
146
+ },
147
+ {
148
+ "step": 26000,
149
+ "val_loss": 4.5980020087132845,
150
+ "wall": 4903.583168268204
151
+ },
152
+ {
153
+ "step": 27000,
154
+ "val_loss": 4.570534922330854,
155
+ "wall": 5264.810336589813
156
+ },
157
+ {
158
+ "step": 28000,
159
+ "val_loss": 4.552568816446366,
160
+ "wall": 5684.886098384857
161
+ },
162
+ {
163
+ "step": 29000,
164
+ "val_loss": 4.54031097933971,
165
+ "wall": 6028.947323799133
166
+ },
167
+ {
168
+ "step": 30000,
169
+ "val_loss": 4.529994341288424,
170
+ "wall": 6374.269299030304
171
+ }
172
+ ],
173
+ "tokens_seen": 3938186976,
174
+ "final_train_loss": 4.545414205116801,
175
+ "final_val_loss": 4.529994341288424,
176
+ "finished": 1786024981.5105116
177
  }