publish trace-repro-lite-ar-s42
Browse files
trace-repro-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-lite-ar-s42/best/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74bfedf46fcaa8eda9c612c5b3732ead9a03da9f78397ed672a0e1f08296266c
|
| 3 |
+
size 179808064
|
trace-repro-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-lite-ar-s42/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c3f07d874b965cdbc61bdee42e4f25e97b86ccbec462634b0f4779cf0ba3b18
|
| 3 |
+
size 179808064
|
trace-repro-lite-ar-s42/trainlog.jsonl
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 1, "train_loss": 4.24583, "val_loss": 5.05005, "epoch_sec": 149.6}
|
| 2 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 2, "train_loss": 3.31301, "val_loss": 4.78943, "epoch_sec": 125.6}
|
| 3 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 3, "train_loss": 2.98457, "val_loss": 4.65473, "epoch_sec": 123.9}
|
| 4 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 4, "train_loss": 2.64557, "val_loss": 4.57144, "epoch_sec": 123.7}
|
| 5 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 5, "train_loss": 2.32419, "val_loss": 4.5358, "epoch_sec": 124.2}
|
| 6 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 6, "train_loss": 2.04471, "val_loss": 4.53471, "epoch_sec": 120.2}
|
| 7 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 7, "train_loss": 1.81486, "val_loss": 4.57338, "epoch_sec": 120.6}
|
| 8 |
+
{"model": "lite", "mode": "ar", "seed": 42, "lr": 0.0002, "canary": false, "init": "scratch", "epoch": 8, "train_loss": 1.6264, "val_loss": 4.57895, "epoch_sec": 120.3}
|