ErrLookup › unslothai/unsloth
unslothai/unsloth
Local UI to run and train LLMs and diffusion models, including Qwen3.8, Kimi K3, MiniMax-H3, Gemma 4, DeepSeek-V4, FLUX and more. · Python · 1,801 source files
Analyzed at 203007d190 on 2026-08-15. 836 documented errors.
| Code / Message | Type | Severity | Tags |
|---|---|---|---|
| The executable at is not stable-diffusion.cpp: its --help… | exception | error | configuration, binary-validation, environment, sd-cpp |
| width and height must be sent together, or both omitted | validation | error | pydantic, validation, video-generation, keyframes, aspect-ratio |
| llama-server does not read an attached value: write | validation | error | llama-server, validation, gnu-syntax, argv, flag-parsing |
| transformer_quant ' ' is unavailable for ' ' : its Modular… | exception | error | video, quantization, validation, minimax-h3, config |
| unavailable for | exception | error | network, huggingface-hub, retry, hf-download, authentication |
| llama-server generation queue is full | exception | warning | llama-server, admission-control, backpressure, concurrency, capacity |
| ' ' is a ')} checkpoint, but it is being loaded against '… | validation | error | diffusion, gguf, flux2, model-loading, validation |
| VirusTotal request failed after | exception | warning | model-loading, security, consent, trust-remote-code, frontend |
| gguf_error | http | error | fastapi, http-409, gguf, llama-server, embeddings, hugging-face, offline |
| ChatGPT authorization is no longer valid. Please reconnect. | exception | error | oauth, authentication, refresh-token, chatgpt, codex |
| mlx-vlm's registered renderer returned an empty prompt. | exception | error | mlx, vlm, chat-template, prompting |
| `penalty` has to be a positive float, but is | validation | error | audio, sampling, repetition-penalty, validation, transformers |
| This checkpoint does not record which optimizer wrote its… | exception | error | training, checkpoint, resume, optimizer, corrupt-file |
| Model path is outside Unsloth storage | http | error | fastapi, unsloth, path-traversal, symlink, security, filesystem, model-deletion |
| MiniMax-H3 needs about | exception | error | system-ram, minimax-h3, video-generation, preflight, diffusers, psutil |
| Linked folder is no longer a regular directory | exception | error | filesystem, symlink, security, sync, toccou |
| VirusTotal returned HTTP | exception | warning | model-loading, transformers, versions, consent, frontend |
| GraphQL failed after | exception | error | github, graphql, network, retry, timeout |
| llama-server does not accept the spaces around | validation | error | llama-server, validation, whitespace, argv, flag-parsing |
| cond_cache_dir is not supported for the sdxl family: its… | http | warning | diffusion, sdxl, cond-cache, config, validation, http-400 |
| Could not update ' ' with the imported example ( ). Nothing… | http | error | filesystem, atomic-write, windows, import, http-409 |
| This quantized (int8/fp8) load was built without LoRA… | validation | error | diffusion, lora, quantization, torchao, model-loading |
| str(exc) | http | warning | fastapi, http-409, presets, quota, validation |
| requires an adapter method (LoRA/QLoRA or Continued… | validation | error | pydantic, lora, fine-tuning, validation |
| The stable-diffusion.cpp binary was replaced by an install… | exception | error | diffusion, sd-cpp, accelerator, stale-state |
| LoRA is not supported for this model/quantisation on the… | validation | error | diffusion, lora, gguf, quantization, torch-compile, engine |
| VirusTotal hash lookup returned a malformed body | exception | error | model-loading, rollback, native-path, file-picker, frontend |
| GGUF variant deletion requires an export directory | http | warning | fastapi, unsloth, gguf, validation, filesystem, model-deletion |
| MLXInferenceBackend cannot load GGUF model | exception | error | gguf, llama-server, huggingface-hub, transient, model-loading |
| MiniMax-H3 reference audio runs up to | validation | error | audio, minimax-h3, upload, duration, validation |
| This is a GGUF model, but the llama.cpp runtime… | exception | error | gguf, llama-cpp, runtime-missing, preflight, huggingface |
| Native path grant timestamps are inconsistent. | validation | error | native-path-lease, validation, timestamps, desktop |
| MiniMax-H3 needs about | exception | error | vram, minimax-h3, video-generation, preflight, diffusers |
| transformer_quant=' ' could not be used: . Choose Auto to… | exception | error | quantization, torchao, gpu, precision, diffusion, host-capability |
| Failed to apply LoRA | validation | error | diffusion, lora, diffusers, error-handling, cleanup |
| ' ' needs diffusers ( ), which this environment cannot… | exception | error | diffusers, environment, dependency, training, install |
| Bad candidate spec fragment | exception | error | oauth, providers, validation, frontend |
| Embedding model ; refusing to load. Set a different RAG… | exception | critical | security, huggingface, pickle, embeddings, model-loading |
| The training checkpoint at | exception | error | checkpoint, resume, filesystem, training |
| Streaming chat-format standardization failed on the first… | validation | error | datasets, streaming, format-conversion, fail-fast |
| extra llama-server args are too long for a Windows command… | validation | error | llama-server, windows, command-line-length, quoting, validation |
| The requested LoRA adapters could not be applied: baking… | exception | error | diffusion, lora, quantization, vram, model-loading |
| Path cannot be empty | http | warning | http-400, linked-folder, path-validation, rag, studio-backend |
| Unsloth MLX: use_adapter must be None, True, False, or a… | validation | error | mlx, lora, type-validation, inference |
| reference audio needs at least one reference image or video… | validation | error | pydantic, validation, video-generation, audio, minimax |
| llama-server embedder failed to become healthy. Last output | exception | critical | subprocess, llama-cpp, embeddings, startup, timeout |
| Chat template exceeds the | validation | warning | validation, chat-template, inference, http-422, size-limit |
| mlx-lm is too old for LRUPromptCache | exception | error | mlx, version-mismatch, prompt-cache, dependencies |
| llama-server flag ' ' requires an integer value | validation | error | llama-server, validation, ctx-size, flag-parsing, configuration |
| Resume checkpoint uses a path from a different operating… | validation | error | path, cross-platform, training-resume, validation |
| Chat template contains unpaired surrogate characters. | validation | error | pydantic, unicode, chat-template, validation |
| Credential or config folders cannot be used for model… | validation | error | huggingface, cache, security, credentials, validation |
| GitHub returned . Token source: . The token is invalid… | exception | critical | github, authentication, http-401, http-403, scopes |
| Cannot apply partial settings patch to corrupt key(s) | exception | error | settings, corruption, transaction, studio-backend |
| family declares no prequant_filenames entry for , so a… | exception | error | http, api, oauth, providers, frontend |
| The local transcription runtime did not answer the request… | exception | error | stt, ggml, network, fallback, engine-health |
| This run has trainable tensors and the checkpoint has : … | validation | error | training, checkpoint, resume, lora, validation |
| Access to ' ' is gated or unauthorized. Accept the model's… | http | error | huggingface, gated-model, auth, token, http-400 |
| ' ' needs diffusers ( ), but this diffusers exports it as a… | exception | error | diffusers, dependency, environment, training, version-mismatch |
| keyframes and references cannot be combined: MiniMax-H3… | validation | error | pydantic, validation, video-generation, keyframes, minimax |
| Assistant message does not match this research run | exception | error | deep-research, precondition, message-binding, state-machine, concurrency |
| Native path grant has expired. | validation | error | native-path-lease, expiry, clock-skew, desktop |
| The Hugging Face cache folder is invalid. | validation | error | huggingface, cache, filesystem, validation, path |
| A prompts list is supported for plain text-to-image only… | validation | error | diffusion, batching, validation, workflow |
| MiniMax-H3 takes at most 12 references in total, got | validation | error | pydantic, validation, video-generation, minimax |
| deadline reached while pacing before | exception | warning | concurrency, model-loading, state, frontend |
| llama-server flag ' ' takes two values | validation | error | llama-server, validation, two-value-flag, argv, flag-parsing |
| Streaming ChatML-to-Alpaca conversion failed on the first… | validation | error | datasets, streaming, format-conversion, alpaca, fail-fast |
| The LoRA selection changed, but a quantized (int8/fp8)… | validation | error | diffusion, lora, quantization, torchao, topology |
| Unknown format_type | validation | error | datasets, format-conversion, validation, configuration |
| apply_chat_template returned None — tokenizer may be… | exception | error | mlx, chat-template, tokenizer, inference |
| (dynamic, from format_error_message: e.g. "Model ' ' not… | exception | error | model-loading, memory, huggingface, generic-wrapper |
| {str(exc)} | http | error | resume, checkpoint, quantization, conflict, http-409 |
| deadline reached before | exception | error | research, approval, state, frontend, version-mismatch |
| ' ' ships LTX-2.3 as single-file checkpoints with no… | validation | error | training, ltx-2, model-loading, diffusers, unsupported |
| Resume checkpoint path could not be resolved. | validation | error | filesystem, training-resume, validation |
| base_precision= needs a CUDA GPU; this host has none. Use… | exception | error | config, gpu, precision, training |
| embedder returned ragged vectors: shape | exception | error | embeddings, numpy, validation, data-integrity |
| GGUF variant ' ' not found in . Available variants | validation | error | gguf, quantization, huggingface, variant-selection, preflight |
| The entire home folder cannot be registered | validation | warning | paths, validation, home-directory, scope-guard |
| base_precision= is not available on this host: torchao is… | exception | error | torchao, rocm, windows, quantization |
| Path under is not allowed | validation | warning | validation, scan-folders, security, path-policy, cross-platform |
| Linked folder no longer resolves to its registered path | exception | error | filesystem, symlink, rag, folder-sync, integrity |
| torch crashes when allocating on | exception | critical | torch, cuda, environment, gpu, embeddings |
| A transformers installation is replacing the latest… | exception | warning | concurrency, transformers, export, retry, studio |
| Decoding is disabled for this feature. Please use… | exception | error | audio, datasets, huggingface, decoding |
| Authorization state did not match. | exception | warning | oauth, csrf, state-mismatch, loopback-server, codex |
| known block type handled by its typed model | validation | info | pydantic, validation, anthropic, content-blocks, internal-invariant |
| the template produced an empty prompt | validation | error | mlx, chat-template, tokenizer, model-loading |
| Image ' ' is too large; maximum is px per side. | http | error | security, decompression-bomb, pillow, image, http-400 |
| local cache path must not contain '..' segments | validation | error | pydantic, validation, filesystem, path-traversal, security |
| RAG is unavailable: the sqlite-vec extension could not be… | http | error | http-503, rag, sqlite-vec, dependency, studio-backend |
| Dataset ' ' must not be a symbolic link. | http | error | security, symlink, path-traversal, fastapi, http-400 |
| ' ' is the outputs folder itself, not a run inside it. Pick… | http | warning | diffusion, output-dir, paths, validation, http-400 |
| dataset_slice_end must be greater than or equal to… | validation | error | pydantic, validation, training, dataset, off-by-one |
| sd-server connection lost during img_gen submit | exception | error | diffusion, sd-cpp, server, network, crash |
| sd-server failed to become ready. Last output | exception | error | diffusion, sd-cpp, server, startup, timeout |
| {_cn_fs.reason} | validation | error | diffusion, security, controlnet, huggingface, validation |
| dataset path must be relative or under a dataset root | validation | error | path-traversal, filesystem, configuration, validation, python |
| extra llama-server args cannot contain a bare value | validation | error | llama-server, validation, argv, flag-parsing |