Upload assets/t75s/loras/summary_t752-teacher-e0400.json with huggingface_hub

assets/t75s/loras/summary_t752-teacher-e0400.json

@@ -0,0 +1,161 @@
+{
+ "char": "yier",
+ "arm": "teacher",
+ "job": "t752",
+ "refs": 2,
+ "buckets": [
+  512
+ ],
+ "steps_cum_target": 400,
+ "base_steps": 0,
+ "save_every": 50,
+ "lr": 0.0003,
+ "rank": 16,
+ "phase": {
+  "pip_s": 27.7,
+  "payload_download_s": 9.6,
+  "cache_extract_s": 11.1,
+  "dit_download_s": 89.2
+ },
+ "gpu": "name, memory.total [MiB], driver_version\nTesla T4, 15360 MiB, 580.178.04\nTesla T4, 15360 MiB, 580.178.04",
+ "cache": {
+  "files": 169,
+  "bytes": 976299608
+ },
+ "n_items": 84,
+ "resume_dir": "",
+ "init_lora": "",
+ "session_steps_planned": 400,
+ "est_s_per_step": 59.0,
+ "train_cmd_tail": [
+  "--num_cpu_threads_per_process",
+  "1",
+  "--mixed_precision",
+  "bf16",
+  "minimax_h3_train_network.py",
+  "--dataset_config",
+  "/tmp/h3/ds/train_r2.toml",
+  "--task",
+  "t2va",
+  "--one_frame",
+  "--video_only",
+  "--dit",
+  "/tmp/h3/models/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
+  "--dit_dtype",
+  "bfloat16",
+  "--mixed_precision",
+  "bf16",
+  "--network_module",
+  "networks.lora_minimax_h3",
+  "--network_dim",
+  "16",
+  "--network_alpha",
+  "16",
+  "--h3_teacher_matching",
+  "--h3_teacher_conditions",
+  "subject_ref",
+  "--h3_teacher_condition_sigma_min",
+  "0.15",
+  "--h3_teacher_condition_sigma_max",
+  "1.0",
+  "--h3_teacher_loss_mag_weight",
+  "0.5",
+  "--h3_teacher_loss_dc_weight",
+  "0.3",
+  "--sdpa",
+  "--split_attn",
+  "--gradient_checkpointing",
+  "--block_swap_h2d_only",
+  "--block_swap_ring_size",
+  "1",
+  "--optimizer_type",
+  "adamw8bit",
+  "--learning_rate",
+  "0.0003",
+  "--lr_scheduler",
+  "constant_with_warmup",
+  "--lr_warmup_steps",
+  "50",
+  "--max_train_steps",
+  "400",
+  "--save_every_n_steps",
+  "50",
+  "--save_precision",
+  "fp32",
+  "--seed",
+  "7",
+  "--save_state",
+  "--output_dir",
+  "/tmp/h3/train_out",
+  "--output_name",
+  "t752-e0000"
+ ],
+ "blocks_to_swap": 40,
+ "train_rc": 0,
+ "train_wall_s": 19494.7,
+ "steps_done_from_log": 400,
+ "s_per_step_measured": 48.74,
+ "loss_first": 0.8,
+ "loss_last": 0.34,
+ "loss_n": 800,
+ "ckpts": [
+  {
+   "cum_steps": 50,
+   "path": "/kaggle/working/out/t752-e0050.safetensors",
+   "bytes": 298260640,
+   "sha256": "b130f03871d9d5737df412dc412d5c9ef6f573b9a89a9ae9c45e76726ccedeee"
+  },
+  {
+   "cum_steps": 100,
+   "path": "/kaggle/working/out/t752-e0100.safetensors",
+   "bytes": 298260648,
+   "sha256": "00f4215fda124a7c7f3146f5ec7de6cfc7f249dd0712fe3404d48f2e59155c2f"
+  },
+  {
+   "cum_steps": 150,
+   "path": "/kaggle/working/out/t752-e0150.safetensors",
+   "bytes": 298260648,
+   "sha256": "a1dbc369763ea166c27548de59c88b1a187670b0b822d438230dfef55f85bfe9"
+  },
+  {
+   "cum_steps": 200,
+   "path": "/kaggle/working/out/t752-e0200.safetensors",
+   "bytes": 298260648,
+   "sha256": "166853e72c5516c33630cb34b67031eb47f118b395a081c0815f7c5f82be0bf4"
+  },
+  {
+   "cum_steps": 250,
+   "path": "/kaggle/working/out/t752-e0250.safetensors",
+   "bytes": 298260648,
+   "sha256": "bce102cc8684530b58199f6929e6beb4bd8330343edb6c07a2bc3fc54f3c7191"
+  },
+  {
+   "cum_steps": 300,
+   "path": "/kaggle/working/out/t752-e0300.safetensors",
+   "bytes": 298260648,
+   "sha256": "c1ab5d23a7ba94700aeebb7f2a4103efa3e503937c4e6031bee5a21d90725002"
+  },
+  {
+   "cum_steps": 350,
+   "path": "/kaggle/working/out/t752-e0350.safetensors",
+   "bytes": 298260640,
+   "sha256": "0c80e6db9785346c34152e7e29429eedadf9958002bb03ba99c6d384a163848d"
+  },
+  {
+   "cum_steps": 400,
+   "path": "/kaggle/working/out/t752-e0400.safetensors",
+   "bytes": 298260648,
+   "sha256": "3307de58ffa0497f96ba4c1719013d4d061d0358b617893135bb0a852443a8e5"
+  }
+ ],
+ "steps_done": 400,
+ "state_tgz": {
+  "path": "/kaggle/working/out/state.tgz",
+  "bytes": 414250703,
+  "cumulative_steps": 400
+ },
+ "peak_vram_mb": 7349,
+ "min_ram_avail_gb": 9.6,
+ "wall_s": 19662.8,
+ "rc": 0
+}