publish trace-repro-todbert-lite-ar-s42
Browse files
trace-repro-todbert-lite-ar-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": 6,
|
| 6 |
+
"num_heads": 8,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "autoregressive",
|
| 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-todbert-lite-ar-s42/best/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:330e89d9d96a987413b48821a9cad0876fe7461dcd789966e9059f2e2ec30512
|
| 3 |
+
size 179808064
|
trace-repro-todbert-lite-ar-s42/config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"input_dim": 768,
|
| 3 |
+
"hidden_dim": 768,
|
| 4 |
+
"output_dim": 768,
|
| 5 |
+
"num_layers": 6,
|
| 6 |
+
"num_heads": 8,
|
| 7 |
+
"dropout": 0.1,
|
| 8 |
+
"max_turns": 64,
|
| 9 |
+
"attention_mode": "autoregressive",
|
| 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-todbert-lite-ar-s42/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:330e89d9d96a987413b48821a9cad0876fe7461dcd789966e9059f2e2ec30512
|
| 3 |
+
size 179808064
|
trace-repro-todbert-lite-ar-s42/trainlog.jsonl
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 1, "train_loss": 4.91352, "val_loss": 5.2449, "epoch_sec": 84.9}
|
| 2 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 2, "train_loss": 4.12681, "val_loss": 5.05074, "epoch_sec": 69.1}
|
| 3 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 3, "train_loss": 3.93202, "val_loss": 4.96235, "epoch_sec": 68.9}
|
| 4 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 4, "train_loss": 3.78034, "val_loss": 4.91769, "epoch_sec": 68.7}
|
| 5 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 5, "train_loss": 3.62823, "val_loss": 4.88299, "epoch_sec": 68.8}
|
| 6 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 6, "train_loss": 3.47093, "val_loss": 4.88183, "epoch_sec": 68.6}
|
| 7 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 7, "train_loss": 3.31467, "val_loss": 4.86332, "epoch_sec": 69.6}
|
| 8 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 8, "train_loss": 3.17237, "val_loss": 4.83914, "epoch_sec": 70.8}
|
| 9 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 9, "train_loss": 3.04632, "val_loss": 4.83368, "epoch_sec": 70.8}
|
| 10 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 10, "train_loss": 2.93813, "val_loss": 4.81707, "epoch_sec": 70.7}
|
| 11 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 11, "train_loss": 2.84382, "val_loss": 4.7945, "epoch_sec": 73.9}
|
| 12 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 12, "train_loss": 2.76378, "val_loss": 4.78936, "epoch_sec": 73.6}
|
| 13 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 13, "train_loss": 2.69561, "val_loss": 4.76946, "epoch_sec": 74.0}
|
| 14 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 14, "train_loss": 2.64153, "val_loss": 4.76156, "epoch_sec": 73.4}
|
| 15 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 15, "train_loss": 2.60807, "val_loss": 4.74569, "epoch_sec": 73.6}
|