indigo
e03cc62548
Wire SCAIL-2 driving + in-context mask conditioning into the trainer via the unified FlexibleStrategy, so the widened patchify_proj (Phase 2) can be trained. - flexible.py: new DrivingConditionConfig (driving-latent concat with a RoPE width offset ΔW) and MaskChannelsConditionConfig (semantic masks -> per-token channels), added to the condition union and get_data_sources. Driving is prepended (cond-first, target stays at the tail for loss slicing); mask channels are written onto the driving tokens via Modality.cond_channels, reusing ltx-core encode_mask_channels. The noisy target keeps a zero mask. - model_loader.load_transformer gains mask_conditioning_channels, widening the video patchify_proj with zero-init columns via a new live-module helper (widen_module_patchify_proj_for_mask_channels). ModelConfig exposes the field. - trainer unfreezes patchify_proj in LoRA mode when mask channels are active (the new input columns are new base params LoRA cannot reach). - configs/scail_animation_lora.yaml plus README / training-modes table rows. Verified on CPU (verify_phase3_trainer.py): config round-trips, prepare_training _inputs builds cond_channels [B,T,56] with the mask on the driving tokens, a tiny widened model forwards and compute_loss returns a finite [B] loss, and the widen helper is output-preserving at zero init. Real training needs Linux+GPU+checkpoint; dataset preprocessing (driving latents + semantic masks) and validation-runner wiring are left for later. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
33 lines
2.9 KiB
Markdown
33 lines
2.9 KiB
Markdown
# Training Configs
|
|
|
|
Example training configurations for the LTX-2 trainer. Each file is a ready-to-run config for one training mode,
|
|
expressed through the unified **flexible** strategy (`name: "flexible"`). Pick the one closest to your use case and
|
|
adjust paths, dataset, and hyperparameters.
|
|
|
|
> 📖 For more information about using each training mode, see [Training Modes Guide](../docs/training-modes.md).
|
|
|
|
## Training Modes
|
|
|
|
| Mode | Video | Audio | Conditions | Config |
|
|
|-----------------------|-----------|-----------|---------------------|--------|
|
|
| **T2V** | Generated | Generated | — | [`t2v_lora.yaml`](./t2v_lora.yaml), [`t2v_lora_low_vram.yaml`](./t2v_lora_low_vram.yaml) (low VRAM) |
|
|
| **I2V** | Generated | Generated | `first_frame` | [`i2v_lora.yaml`](./i2v_lora.yaml) |
|
|
| **Video Extension** | Generated | Generated | `prefix`/`suffix` | [`video_extend_lora.yaml`](./video_extend_lora.yaml) (forward), [`video_suffix_lora.yaml`](./video_suffix_lora.yaml) (backward) |
|
|
| **V2V IC-LoRA** | Generated | — | `reference` | [`v2v_ic_lora.yaml`](./v2v_ic_lora.yaml) |
|
|
| **A2V** | Generated | Frozen | — | [`a2v_lora.yaml`](./a2v_lora.yaml) |
|
|
| **V2A (Foley)** | Frozen | Generated | — | [`v2a_lora.yaml`](./v2a_lora.yaml) |
|
|
| **Video Inpainting** | Generated | — | `mask` | [`video_inpainting_lora.yaml`](./video_inpainting_lora.yaml) |
|
|
| **Video Outpainting** | Generated | — | `spatial_crop` | [`video_outpainting_lora.yaml`](./video_outpainting_lora.yaml) |
|
|
| **T2A** | — | Generated | — | [`t2a_lora.yaml`](./t2a_lora.yaml) |
|
|
| **Audio Extension** | — | Generated | `prefix`/`suffix` | [`audio_extend_lora.yaml`](./audio_extend_lora.yaml) (forward), [`audio_suffix_lora.yaml`](./audio_suffix_lora.yaml) (backward) |
|
|
| **Audio Inpainting** | — | Generated | `mask` | [`audio_inpainting_lora.yaml`](./audio_inpainting_lora.yaml) |
|
|
| **A2A IC-LoRA** | — | Generated | `reference` | [`a2a_ic_lora.yaml`](./a2a_ic_lora.yaml) |
|
|
| **AV2AV IC-LoRA** | Generated | Generated | `reference` (both) | [`av2av_ic_lora.yaml`](./av2av_ic_lora.yaml) |
|
|
| **SCAIL Animation** | Generated | — | `driving` + `mask_channels` | [`scail_animation_lora.yaml`](./scail_animation_lora.yaml) |
|
|
|
|
The [`accelerate/`](./accelerate) directory holds the Accelerate launch configs (FSDP, DDP) for multi-GPU training.
|
|
|
|
> **SCAIL Animation** (SCAIL-2 character animation) also sets `model.mask_conditioning_channels: 56` to widen the
|
|
> video `patchify_proj` for the in-context mask channels; in LoRA mode that projection is unfrozen so the new columns
|
|
> train. See [Training Modes Guide](../docs/training-modes.md).
|