publish trace-repro-deep-bidi-s123
Browse files
trace-repro-deep-bidi-s123/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-s123/best/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a24f30249a3e1a83488151096d82c4e30aefe328f06de1bcc5eb7f033d14fba4
|
| 3 |
+
size 349928328
|
trace-repro-deep-bidi-s123/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-s123/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:654942c0a8254c6ccf0c052bce693273ec61f6ba4b27201958161a77ac8d9604
|
| 3 |
+
size 349928328
|
trace-repro-deep-bidi-s123/trainlog.jsonl
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 1, "train_loss": 3.50183, "val_loss": 4.06671, "epoch_sec": 203.7}
|
| 2 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 2, "train_loss": 2.49594, "val_loss": 3.56176, "epoch_sec": 196.5}
|
| 3 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 3, "train_loss": 1.96818, "val_loss": 3.28322, "epoch_sec": 200.1}
|
| 4 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 4, "train_loss": 1.55854, "val_loss": 3.18526, "epoch_sec": 202.3}
|
| 5 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 5, "train_loss": 1.23335, "val_loss": 3.15148, "epoch_sec": 202.8}
|
| 6 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 6, "train_loss": 0.97875, "val_loss": 3.158, "epoch_sec": 200.6}
|
| 7 |
+
{"model": "deep", "mode": "bidi", "seed": 123, "lr": 0.0001, "canary": false, "init": "scratch", "epoch": 7, "train_loss": 0.78172, "val_loss": 3.15297, "epoch_sec": 196.7}
|