Upload config.json
Browse files
CurriculumStages/0_228M/config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"AdaLN_Timestep_Embed_Dim": 256,
|
| 3 |
+
"adafactor_rel_step": true,
|
| 4 |
+
"adafactor_warmup_init": true,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LexaLCM"
|
| 7 |
+
],
|
| 8 |
+
"batch_size": 16,
|
| 9 |
+
"bf16": true,
|
| 10 |
+
"cfg_scale": 0.0,
|
| 11 |
+
"clip_threshold": 1.0,
|
| 12 |
+
"contextualizer": 1,
|
| 13 |
+
"custom_multi_gpu": true,
|
| 14 |
+
"d_ff": 8192,
|
| 15 |
+
"d_latent": 1024,
|
| 16 |
+
"d_model": 2048,
|
| 17 |
+
"denoiser": 0,
|
| 18 |
+
"denoiser_iterations_inference": 40,
|
| 19 |
+
"denoiser_iterations_pretrain": 100,
|
| 20 |
+
"dropout_context": 0.1,
|
| 21 |
+
"dropout_denoiser": 0.15,
|
| 22 |
+
"dropout_latent": 0.1,
|
| 23 |
+
"eval_every": 0,
|
| 24 |
+
"gpus": {
|
| 25 |
+
"contextualizer": 1,
|
| 26 |
+
"custom_multi_gpu": true,
|
| 27 |
+
"denoiser": 0,
|
| 28 |
+
"other": 1
|
| 29 |
+
},
|
| 30 |
+
"grad_norm_log_every": 20,
|
| 31 |
+
"input_dim": 1024,
|
| 32 |
+
"learning_rate": "None",
|
| 33 |
+
"max_grad_norm": "None",
|
| 34 |
+
"max_seq_len": 64,
|
| 35 |
+
"max_steps": 250000,
|
| 36 |
+
"model_type": "lexa_lcm_pre3",
|
| 37 |
+
"n_heads": 32,
|
| 38 |
+
"num_context_layers": 5,
|
| 39 |
+
"num_denoiser_layers": 15,
|
| 40 |
+
"num_denoising_steps": 100,
|
| 41 |
+
"num_workers": 20,
|
| 42 |
+
"optimizer": "adafactor",
|
| 43 |
+
"other": 1,
|
| 44 |
+
"output_dir": "./outputs",
|
| 45 |
+
"resume_from": "None",
|
| 46 |
+
"save_every": 5000,
|
| 47 |
+
"torch_dtype": "float32",
|
| 48 |
+
"transformers_version": "4.51.3",
|
| 49 |
+
"warmup_steps": 500,
|
| 50 |
+
"weight_decay": 0.01
|
| 51 |
+
}
|