unslothai/unsloth
Documented errors, page 2 of 9. Back to unslothai/unsloth
| Code / Message | Type | Severity | Tags |
|---|---|---|---|
| ' ' is still cancelling; try again in a moment. | exception | warning | stt, whisper, ggml, download, cancellation, race |
| Refusing to train from untrusted base model | validation | error | training, security, model-loading, trust, pickle-risk |
| sd-server is not running. | exception | error | diffusion, sd-cpp, server, lifecycle |
| Native path grant is required. | http | error | http-400, native-path-lease, file-attach, authorization, rag, studio-backend |
| Unsloth MLX: cannot disable adapter layers without their… | exception | error | mlx, lora, model-architecture, inference |
| Could not install mamba-ssm, required by this Mamba model. | exception | error | mamba, dependency-install, gpu, model-loading, python |
| hf_dataset is too long (max 256 chars) | validation | error | pydantic, validation, training, huggingface, dataset |
| text_encoder_quant=' ' could not be used: . Leave it unset… | http | error | video, quantization, offload, hardware-support |
| Training images cannot be changed while diffusion training… | http | warning | diffusion, dataset, interlock, mutation, http-409 |
| base_precision= needs a dense base repo, but ' ' is already… | validation | error | training, quantization, base-model, configuration, validation |
| too many extra llama-server args | validation | error | llama-server, validation, argv, configuration |
| Expected a number, got a boolean. | validation | warning | pydantic, validation, http-422, boolean-coercion, llm-config |
| carries s of audio for a s clip. MiniMax-H3 denoises video… | exception | error | audio, video, dataset, minimax-h3 |
| Seed file does not exist | validation | error | pydantic, config-validation, filesystem, data-designer |
| Dataset upload too large. Maximum is | http | error | upload, http-413, fastapi, size-limit |
| Reference images are not supported for the | validation | error | diffusion, validation, reference-images, model-family |
| Failed to decrypt API key. The server public key may have… | http | error | encryption, rsa, api-key, provider-credentials |
| dataset_streaming requires a plain split name in | validation | error | pydantic, huggingface, datasets, streaming, validation |
| load error | exception | warning | audio, attachments, validation, frontend |
| This execution artifact is outside the Recipe Studio… | exception | error | validation, path-traversal, huggingface, data-recipe, studio |
| ControlNet is not supported for the | validation | error | diffusion, controlnet, model-family, validation |
| fill_absent_fields cannot be combined with remove. | exception | warning | fastapi, http-409, validation, mutually-exclusive, model-overrides |
| Dropped dataset could not be read. | http | error | upload, native-import, filesystem, http-400 |
| [ERROR] Unsloth frontend build not found. Tried | console | critical | startup, frontend-build, path-shadowing, installation, cli |
| Network paths are not supported for native grants. | exception | error | native-path-lease, windows, unc, network-path |
| s3_config requires either use_iam_role=True or both… | validation | error | pydantic, validation, s3, credentials, aws, training |
| The stable-diffusion.cpp server binary was replaced by an… | exception | error | concurrency, race-condition, retryable, accelerator, sd-cpp |
| llama-server embedder POST | exception | error | http, embeddings, llama-cpp, batching |
| Metadata unavailable while resolving GGUF variant | exception | error | network, huggingface, gguf, download, retry |
| ' ' cannot be loaded | validation | error | model-validation, unsupported-model, edit-models, diffusion, family-detection |
| Unsupported attention_backend | validation | error | diffusion, attention-backend, validation, configuration |
| GGUF LoRA adapters are not supported on the diffusers engine | validation | error | diffusion, lora, gguf, safetensors, validation |
| Native path grant issue time is invalid. | validation | error | native-path-lease, clock-skew, validation, desktop |
| Folder was concurrently removed | validation | warning | race-condition, sqlite, scan-folders, concurrency |
| save_directory may not contain '..' segments | validation | error | validation, security, path-traversal, export, http-422 |
| has cached pickle weights that cannot be security-scanned… | http | error | fastapi, http-403, security, hugging-face, pickle, offline, embeddings |
| ${parseErrorText(response.status, body)} | http | error | settings, retry-queue, http, schema-drift |
| ema_decay must be a number, got | validation | error | training, ema, configuration, validation, type-coercion |
| Failed to export base model | http | error | export, http-500, fastapi, unsloth |
| pandas is required for unstructured seed processing | exception | error | pandas, dependency, import-error, plugin |
| Provider base URL hostname could not be resolved. | validation | error | network, dns, validation, provider-config, ssrf |
| The source run contains invalid training configuration and… | http | error | training, resume, validation, schema-mismatch, http-409 |
| Unknown completion masking template | validation | error | training, completion-masking, configuration, validation |
| is not supported for the ' ' model family. | validation | error | diffusion, workflow, model-family, validation |
| model_path is required | http | warning | fastapi, unsloth, validation, request-body, model-deletion |
| Resume checkpoint must be inside Unsloth outputs. | validation | error | path, sandboxing, training-resume, validation |
| The ChatGPT account identifier was missing. | exception | error | oauth, jwt, claims, token-validation, codex |
| This LTX checkpoint stores scaled fp8 weights, which this… | exception | critical | video, ltx, quantization, checkpoint, model-loading |
| Could not infer role/content keys for chat column | validation | error | datasets, format-conversion, chatml, data-quality |
| extra llama-server args cannot contain unpaired surrogate… | validation | error | llama-server, unicode, surrogate, validation, encoding |
| No valid examples after Whisper preprocessing | validation | error | audio, whisper, asr, dataset, data-quality |
| Subprocess error | exception | error | inference, subprocess, runtime-error, relay |
| Unknown provider type | http | error | providers, registry, connectivity, validation |
| base_precision='int8' needs a functional torchao install… | exception | error | dependencies, torchao, quantization, training |
| learning_rate is required | validation | error | pydantic, validation, training, hyperparameters |
| Refusing to delete storage root | http | error | fastapi, unsloth, destructive-operation, guard, filesystem, model-deletion |
| Authorization callback was invalid or already used. | exception | warning | oauth, consent-denied, replay, loopback-server, codex |
| HTTP ( ) | http | error | openai, containers, encryption, external-api |
| Image ' ' is too large ( x ); maximum is px per side. | http | error | image, pillow, http-400, validation, dataset |
| local cache path is too long (max 4096 chars) | validation | error | pydantic, validation, training, filesystem, paths |
| ' ' is not a loadable single-file checkpoint (expected a… | validation | error | validation, safetensors, single-file, file-format, diffusion |
| Streaming Alpaca-to-ChatML conversion failed on the first… | validation | error | datasets, streaming, format-conversion, chatml, fail-fast |
| WebM export failed (libvpx-vp9 unavailable?) | exception | error | video, webm, codec, libvpx, dependency |
| ChatGPT returned an invalid access token. | exception | error | oauth, jwt, token-validation, codex |
| Device authorization failed. Enable device-code login in… | exception | error | oauth, device-flow, chatgpt, codex, account-settings |
| embedder returned vectors for inputs | exception | error | embeddings, protocol, validation, batching |
| Invalid model snapshot repository ID. | http | error | models, huggingface, validation, repo-id, api |
| learning_rate must be a number, not a bool | validation | error | pydantic, validation, training, yaml, type-coercion |
| This checkpoint's optimizer state was written by | exception | error | training, checkpoint, resume, optimizer, environment |
| Unsloth MLX: max_grad_value= | validation | error | training, mlx, config-validation, gradient-clipping |
| A diffusion load is already in progress. | exception | error | concurrency, load-state, diffusion, locking, model-loading |
| Dataset ' ' escapes the Studio datasets directory. | http | error | security, path-traversal, symlink, fastapi, http-400 |
| Native path grant contains invalid characters. | validation | error | native-path-lease, injection, validation, payload |
| Failed to start | http | critical | subprocess, spawn, environment, http-500, download |
| Model ' ' has no chat_template set in its… | validation | error | chat-template, tokenizer, huggingface, model-config |
| Failed to delete dataset from | http | error | cache, filesystem, permissions, partial-failure, http-500 |
| No audio column found in dataset. Columns | validation | error | audio, dataset, csm, speech, validation |
| Provider base URL points at a private address, which is… | validation | error | network, ssrf, validation, provider-config, security |
| Authorization flow was cancelled before credentials were… | exception | error | oauth, race-condition, cancellation, persistence, codex |
| installer timed out after | exception | error | timeout, installer, subprocess, network, python |
| S3 prefix contains mixed dataset formats | validation | error | s3, dataset, validation, data-format |
| The managed RAG uploads folder cannot be linked | validation | warning | paths, rag, validation, recursion-guard |
| duplicate LoRA id ' '; list each adapter at most once | validation | error | pydantic, validation, lora, diffusion, quantization, duplicates |
| Upscale would not enlarge this image: its longest side | validation | error | diffusion, upscale, validation, image-size |
| Local pipeline directory has no model_index.json | validation | error | local-path, pipeline, model-index, validation, diffusion |
| Non-GGUF diffusion loads are restricted to unsloth/* repos… | validation | error | security, trust-gate, unsloth, huggingface, model-validation, diffusion |
| ' ' is a single-file GGUF repo; load it with model_kind… | validation | error | gguf, huggingface, pipeline, validation, load-kinds, diffusion |
| An LLM training job is already running. Stop it before… | http | warning | diffusion, llm, gpu, interlock, conflict, http-409 |
| base_precision='mxfp8' needs a Blackwell (sm100+) GPU; this… | exception | error | gpu, blackwell, quantization, mxfp8 |
| Diffusion generation was cancelled. | exception | info | diffusion, cancellation, concurrency, controlnet |
| : refusing to pad a quantized Linear whose activation… | exception | error | quantization, numerics, validation, padding |
| GGUF conversion produced a symlink, refusing to relocate it | exception | error | export, gguf, filesystem, symlink, studio |
| installer exited | exception | error | installer, subprocess, update, python |
| : not valid JSON | exception | error | sse, streaming, fetch, research, frontend |
| backend switch must preserve | exception | error | llama-cpp, runtime-installer, backend-switch, release-pin, invariant |
| Local commands aren't enabled on this server. To allow… | http | warning | mcp, stdio, security, http-400, config |
| ChatGPT credential update is busy. Please retry. | exception | warning | oauth, file-lock, concurrency, retry, codex |
| hf_dataset contains invalid characters or path segments | validation | error | pydantic, validation, training, huggingface, dataset, sanitization |
| Invalid or expired video link. | http | error | http-401, hmac, signed-url, video-streaming, fastapi |
| Cannot start diffusion (Images) training over the API while… | http | warning | diffusion, api-key, gpu, inference, conflict, http-409 |