publish trace-repro-deep-bidi-s42
Browse files
trace-repro-deep-bidi-s42/best/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"input_dim": 768,
|
| 3 |
+
"hidden_dim": 768,
|
| 4 |
+
"output_dim": 768,
|
| 5 |
+
"num_layers": 12,
|
| 6 |
+
"num_heads": 12,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "bidirectional",
|
| 10 |
+
"use_speaker_embeddings": true,
|
| 11 |
+
"num_speakers": 4,
|
| 12 |
+
"layer_norm": true,
|
| 13 |
+
"ff_dim": 3072,
|
| 14 |
+
"activation": "gelu",
|
| 15 |
+
"output_residual": false,
|
| 16 |
+
"arch": "v2",
|
| 17 |
+
"head_mode": "tied_continuous",
|
| 18 |
+
"head_transform": true
|
| 19 |
+
}
|
trace-repro-deep-bidi-s42/best/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63e36ce9ae807cbd153fead58a0376601fe323be29f0e43c910cc789d2dd4dff
|
| 3 |
+
size 349928328
|
trace-repro-deep-bidi-s42/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"input_dim": 768,
|
| 3 |
+
"hidden_dim": 768,
|
| 4 |
+
"output_dim": 768,
|
| 5 |
+
"num_layers": 12,
|
| 6 |
+
"num_heads": 12,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "bidirectional",
|
| 10 |
+
"use_speaker_embeddings": true,
|
| 11 |
+
"num_speakers": 4,
|
| 12 |
+
"layer_norm": true,
|
| 13 |
+
"ff_dim": 3072,
|
| 14 |
+
"activation": "gelu",
|
| 15 |
+
"output_residual": false,
|
| 16 |
+
"arch": "v2",
|
| 17 |
+
"head_mode": "tied_continuous",
|
| 18 |
+
"head_transform": true
|
| 19 |
+
}
|
trace-repro-deep-bidi-s42/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6216dd7446359cca1402d5840d969a04f7484944308f2de83c2ef5ae9abcded
|
| 3 |
+
size 349928328
|
trace-repro-deep-bidi-s42/trainlog.jsonl
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 1, "train_loss": 3.49699, "val_loss": 4.04271, "epoch_sec": 211.5}
|
| 2 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 2, "train_loss": 2.49338, "val_loss": 3.55685, "epoch_sec": 199.2}
|
| 3 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 3, "train_loss": 1.97394, "val_loss": 3.28957, "epoch_sec": 199.2}
|
| 4 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 4, "train_loss": 1.5699, "val_loss": 3.19654, "epoch_sec": 199.3}
|
| 5 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 5, "train_loss": 1.24134, "val_loss": 3.16813, "epoch_sec": 199.3}
|
| 6 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 6, "train_loss": 0.98636, "val_loss": 3.16617, "epoch_sec": 198.9}
|
| 7 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 7, "train_loss": 0.78816, "val_loss": 3.19226, "epoch_sec": 199.2}
|
| 8 |
+
{"model": "deep", "mode": "bidi", "seed": 42, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 8, "train_loss": 0.63517, "val_loss": 3.19431, "epoch_sec": 199.4}
|