Skip to content
vision-sft ยท wshobson-agents ยท Flowy
Fine-tune vision-language models (VLMs) with supervised learning on image+text data. Use when adapting a VLM to a visual domain or task, configuring frozen-vision-tower LoRA, or debugging a VLM fine-tune that trains without learning.
One skill from wshobson-agents .
$ npx -y skills add wshobson/agents --skill vision-sft --agent claude-codeInstalls just this skill. Get the whole plugin for auto-invocation.
โก How it fires
How this skill gets triggered: by you, by Claude, or both.
Fires itself Claude auto-loads it when your prompt matches the work. You can call it Invoke it directly when you want it. Slash command /vision-sft๐๏ธ Context preview
The summary Claude sees to decide when to auto-load this skill.
Fine-tune vision-language models (VLMs) with supervised learning on image+text data. Use when adapting a VLM to a visual domain or task, configuring frozen-vision-tower LoRA, or debugging a VLM fine-tune that trains without learning. ๐ฆ Ships with wshobson-agents
</> SKILL.md
vision-sft.SKILL.md
---
name : vision-sft
description : Fine-tune vision-language models (VLMs) with supervised learning on image+text data. Use when adapting a VLM to a visual domain or task, configuring frozen-vision-tower LoRA, or debugging a VLM fine-tune that trains without learning.
---
# Vision-Language SFT
This skill assumes `finetuning-method-selection`
already routed here: the data shape is
image+text demonstrations, not preference pairs
or a verifiable reward signal, and the base is a
vision-language model rather than a text-only
one. `lora-qlora-recipes` covers the text-only
LoRA/QLoRA recipe this skill specializes for the
vision tower and projector; read that skill first
if the LoRA fundamentals (rank, alpha, target
modules) aren't already familiar.
**Input:** an image+text dataset and a VLM base
model already picked from the model catalog.
**Output format:** a validated adapter config โ
which components are frozen, LoRA target modules,
and a `min_pixels`/`max_pixels` budget โ that
`llm-finetuning-training-engineer` consumes
directly when it generates a runnable script.
## Quick Reference
| Situation | Default |
|---|---|
| Adapting behavior on familiar images | Frozen tower+projector, LoRA r=8โ16, ฮฑ=16โ32 |
Flowy AI Flows that just work. Hand-picked Claude Code plugins that fire the right skill as you prompt.
| Visual domain shift | Unfreeze last-6 ViT layers, vision LR 5โ10x lower |
| Doesn't fit in bf16 at target rank | QLoRA โ frozen vision tower only |
| `fast_inference=True` | `finetune_vision_layers=False` |
| Loss normal, eval not improving | Check the Two Silent Killers below first |
## The Consensus Recipe
Freeze the vision tower and the projector. Put
LoRA on the LLM only, all-linear (the same
attention + MLP target list as text-only SFT โ
see `lora-qlora-recipes`), at **r=8โ16,
ฮฑ=16โ32**. This is the settled default for
adapting a VLM's behavior without disturbing how
it sees.
- **The vision tower and projector stay frozen by
default.** They already encode a general visual
representation; retraining them is rarely
necessary and adds risk without adding
capability for most tasks.
- **LoRA rank runs lower than the text-only
general default** (r=8โ16 here vs r=16โ32 for
text-only SFT) because the LLM-only adapter is
adapting behavior, not injecting new visual
knowledge.
- **QLoRA is permitted only with a frozen vision
tower.** Quantizing the base while also
unfreezing and training vision layers is
unsupported and unstable โ treat this as a hard
pairing rule, not a tunable. If the vision tower
needs to unfreeze, drop QLoRA and use bf16 LoRA
instead.
```python
# freeze tower + projector; LoRA on LLM only
for name, param in model.named_parameters():
if "vision_tower" in name or "projector" in name:
param.requires_grad = False
target_modules = [
"q_proj", "k_proj", "v_proj", "o_proj",
"gate_proj", "up_proj", "down_proj",
] # LLM-only, all-linear โ r=8-16, alpha=16-32
```
## When to Unfreeze
Unfreezing vision layers is a deliberate
escalation, not a default decision โ reach
for it only when the domain shift is
visual, not textual.
- **Unfreeze only for visual domain
shift.** If the task is teaching new
behavior on images the tower already
understands (charts, everyday photos),
the frozen-tower recipe above is
sufficient. Unfreeze when the visual
domain itself is unfamiliar to the
tower โ satellite imagery, medical
scans, dense technical diagrams โ and
the frozen-tower recipe plateaus.
- **Last-6 ViT layers is the sweet
spot.** Unfreezing the final six
vision-transformer layers (not the
whole tower) measured **+1.7pt DocVQA
at ~1.75x training cost** over the
frozen baseline. Treat six layers as
the ceiling worth paying for; going
further spends compute without a
matched result.
- **Vision LR must run 5โ10x lower than
the LLM LR when unfrozen.** The vision
tower's pretrained representation is
more fragile than the LLM's adapter;
the same LR for both risks overwriting
the visual representation faster than
the LLM adapter can compensate.
- **High LoRA rank on the patch-
embedding layer risks NaN.** If patch
embedding is in the unfrozen set, keep
its rank low and watch early-step loss
closely โ one of the most fragile
places to apply LoRA in a VLM.
## The Two Silent Killers
Both produce a run that trains without error and
without learning: the loss curve looks normal,
the model doesn't improve, and neither throws an
exception โ both need an explicit pre-training
check, not just a clean training log.
- **Image-tag/count mismatch.** Every image
placeholder token in the templated text must
map 1:1 to a media item actually passed to the
collator. A mismatch (one placeholder, zero or
two images attached; or an image with no
placeholder) doesn't error in most collators โ
it silently misaligns image and text, and the
model "trains but learns nothing." Validate the
1:1 placeholder-to-media mapping before training
starts, on every example, not just a sample.
Full validation-checklist detail:
`references/collators-and-pitfalls.md`.
- **`min_pixels`/`max_pixels` resolution budget.**
This pair is the single most consequential
hyperparameter for quality and memory in VLM
SFT โ more than rank, alpha, or LR. Too low
silently downsamples images below what the task
needs (small document text becomes unreadable
even though training "succeeds"); too high blows
the activation memory budget or forces too small
a batch to train stably. Set it deliberately per
dataset, don't leave it at a framework default.
## Unsloth Specifics
- **`UnslothVisionDataCollator`** is the collator
Unsloth expects for VLM SFT โ it handles the
image-tag alignment and per-architecture
processor contract described in
`references/collators-and-pitfalls.md`. Don't
substitute a text-only collator for VLM data.
- **`finetune_vision_layers=False` is required
when `fast_inference=True`.** vLLM cannot serve
LoRA adapters on vision layers, so a fast-
inference setup that also unfreezes vision
layers fails at serve time even if training
succeeds. If the recipe calls for unfreezing the
last-6 ViT layers (see When to Unfreeze above),
fast inference is off the table for that run โ
choose one or the other, not both.
## Model Choice
Base VLM choice is out of scope for this skill โ
it lives in one place, the model catalog at
`finetuning-method-selection`'s
`references/model-catalog.md`. This skill and its
references describe recipes by architecture
family only, never by recommending one model over
another.
VLM reinforcement learning (VLM-GRPO) is
reference-only in this plugin โ the fragmented
tooling and reward-hacking failure modes specific
to VLM-RL are covered in `grpo-rlvr-training`,
not here. This skill's scope stops at supervised
fine-tuning.
## Failure Modes
The recurring mistake across every section above
is treating a clean loss curve as proof the run
is healthy. A normal-looking curve is consistent
with **both** a working run **and** either silent
killer, since the model trains on *something*
either way โ just not the aligned image-text
signal when a killer is present. A flat eval score
next to a normal loss curve means re-run the
checklist in `references/collators-and-pitfalls.md`
before touching any hyperparameter.
## References
- `references/collators-and-pitfalls.md` โ per-
architecture collator table, dataset-format
examples with image placeholders, a pre-
training validation checklist, and the two-
stage projector-alignment recipe as an advanced
pattern.
Related skills: `finetuning-method-selection`
routes here; `lora-qlora-recipes` covers the
text-only LoRA fundamentals this skill
specializes; `grpo-rlvr-training` covers VLM-RL
(reference-only); `dataset-curation` covers
image+text dataset preparation this skill doesn't.
ยฉ 2026 Flowy ยท Free and open source
Built for Claude Code ยท Not affiliated with Anthropic
