unslothai/unsloth
Documented errors, page 8 of 9. Back to unslothai/unsloth
| Code / Message | Type | Severity | Tags |
|---|---|---|---|
| Studio cannot use this cache folder | validation | error | huggingface, cache, disk-full, io-error, filesystem |
| The previous dictation worker did not exit and still holds… | http | error | stt, worker, process, memory, busy, gpu |
| Image not found. | http | error | http-404, fastapi, dataset, image |
| No text column found in dataset. Columns | validation | error | audio, dataset, csm, speech, validation |
| ChatGPT subscription routing is fixed. | http | error | chatgpt, oauth, base-url, validation |
| Invalid gpu_ids : GPU IDs must be non-negative. Rejected… | validation | error | gpu, vulkan, hardware, validation |
| Native path is no longer a regular file. | validation | error | symlink, security, toctou, filesystem |
| Could not decode image | validation | error | diffusion, image, pillow, decode, client-error |
| learning_rate must be < 1.0 | validation | error | pydantic, validation, training, hyperparameters, divergence |
| Stream response missing body | http | error | streaming, sse, fetch, invariant |
| save_total_limit must be >= 0 (0 keeps every checkpoint) | validation | error | training, checkpointing, configuration, validation |
| Unsupported transformer_quant | validation | error | validation, quantization, transformer, api |
| Another dictation model | http | warning | stt, mtmd, download, concurrency, one-at-a-time |
| dataset_name must be a Hugging Face repo id like org/repo | http | error | huggingface, validation, http-400, dataset |
| File exceeds the MB upload limit. | http | warning | http-413, file-upload, size-limit, rag, studio-backend |
| Keep KV on idle unload must be true or false. | validation | error | settings, validation, boolean-coercion |
| Not an image file. Allowed | http | error | validation, file-type, fastapi, http-400 |
| This video has no captions. | exception | warning | youtube, transcript, no-captions, data-availability |
| Unsupported file type | http | warning | http-400, file-upload, extension-allowlist, rag, studio-backend |
| YouTube returned unreadable caption data. | exception | error | youtube, transcript, json-parse, network |
| Expected numbers, got a boolean. | validation | warning | pydantic, validation, http-422, boolean-coercion, gpu-ids, lists |
| in_features is not divisible by the ConvRot group | exception | error | convrot, shape-mismatch, linear-layer, quantization |
| seed must fit in torch's 64-bit range | validation | error | training, seed, pytorch, validation, configuration |
| activation rotation target | exception | error | convrot, double-apply, state-mismatch, quantization |
| Dropped dataset is empty | http | error | upload, native-import, empty-file, http-400 |
| must be a multiple of 16 | validation | error | pydantic, validation, diffusion, image-dimensions, alignment |
| Total upload limit ( ) exceeded | http | error | upload, quota, http-413, unstructured |
| ControlNet is not yet supported on the native sd.cpp… | validation | error | capability, validation, controlnet, sd-cpp |
| Local checkpoint ' ' is not a .gguf file; a 'gguf' load… | exception | error | video, local-path, gguf, file-extension, validation |
| STT model ' ' (GGUF) is not downloaded. Download it in… | exception | error | stt, ggml, model-download, configuration |
| Unsupported unstructured seed file type | exception | error | file-type, validation, plugin, chunking |
| cannot rotate : not an nn.Linear on this model | exception | error | convrot, fqn, model-mismatch, quantization |
| GIF export needs the 'av' and 'Pillow' packages. | exception | error | video, gif, dependency, pyav, pillow |
| Password change required | http | error | authorization, password, http-403, studio |
| use_oauth must be true or false | http | warning | mcp, validation, http-400, oauth |
| WebM export failed: the clip has no video stream. | exception | error | video, webm, corrupt-file, validation |
| A reference is too large | exception | error | video, reference-media, size-limit, payload |
| every seed must be between 0 and 2**53 - 1 (JSON-safe) | validation | error | diffusion, seeds, validation, json-safety |
| invalid base64 payload | http | error | base64, upload, validation, http-400 |
| ' ' is not a supported diffusion image model. Supported… | validation | error | validation, model-support, family-detection |
| cfg_dropout must be a number, got | validation | error | training, config, validation, cfg |
| data_dir is not a directory | exception | error | filesystem, dataset, video, training |
| unsupported file type | http | error | file-type, upload, validation, http-422 |
| Execution artifacts are no longer available. | exception | error | filesystem, huggingface, data-recipe, studio |
| filename is required when using content_base64 | validation | warning | validation, pydantic, api, http-422, upload |
| is_enabled must be true or false | http | warning | mcp, validation, http-400, boolean |
| That reference file carries no audio track. | validation | error | audio, minimax-h3, upload, validation, pyav |
| is an image-editing model and does not support masks… | validation | error | diffusion, validation, inpainting, mask, image-editing |
| ChatGPT returned a malformed stream. | exception | error | sse, streaming, codex, json, parsing |
| The dictation model changed while this recording was being… | http | warning | stt, model-switch, concurrency, race-condition |
| failed to delete uploaded files | http | error | filesystem, delete, http-500, race-condition |
| file_ids must not be empty | validation | warning | validation, pydantic, api, http-422, upload |
| Speech-to-text needs PyTorch, Transformers, and PyAV. Run… | http | error | stt, dependencies, environment, import-error |
| The pinned archive ended unexpectedly | exception | error | tarfile, truncated-download, corrupt-archive, extraction |
| Could not decode the audio. | exception | error | stt, ggml, response-schema, audio-decode, http |
| file_names must be provided and same length as file_ids | validation | warning | validation, pydantic, api, http-422, upload |
| Media auto-unload idle seconds must be 0 (off) or at least | validation | error | settings, validation, range-check |
| No images found in ' '. | http | error | huggingface, upstream, import, http-502, empty-data |
| Failed to export merged model | http | error | export, http-500, merge, logging |
| flow_shift must be a positive number or 'auto', got | validation | error | training, config, flow-matching, validation, scheduler |
| MCP server not found | http | error | mcp, http-404, crud, stale-id |
| Failed to load checkpoint | http | error | export, http-500, logging, checkpoint |
| {validation_message} | http | error | training, resume, path-validation, http-400 |
| Empty upload payload | http | error | upload, empty-file, multipart, http-400 |
| Speech-to-text needs the PyAV package to decode audio. Run… | http | error | stt, dependency, pyav, ffmpeg, import |
| Native path grant has an invalid format. | validation | error | type-error, api-contract, validation |
| No dataset file was provided | http | error | upload, multipart, validation, http-400 |
| Uploaded file is empty. | http | warning | http-400, file-upload, empty-file, rag, studio-backend |
| Unknown example dataset | http | error | http-404, fastapi, dataset, examples |
| This video's captions are empty. | exception | warning | youtube, transcript, empty-response |
| save_directory may not contain control characters | validation | error | validation, security, crlf, export, http-422 |
| The pinned archive has too many entries | exception | error | tarfile, resource-limit, extraction, security |
| UNSLOTH_STUDIO_MCP_TOKEN is required when MCP is enabled | exception | critical | mcp, auth, env, security, startup |
| activation rotation target | exception | error | convrot, ordering, fqn, quantization |
| Do not reserve system RAM must be true or false. | validation | error | settings, boolean-coercion, ram, api-validation |
| Invalid gpu_ids : GPU IDs must be non-negative. Rejected… | validation | error | gpu, cuda, hardware, validation |
| Delete failed ( ) | http | error | prompts, delete, http, concurrency |
| Folder name cannot be empty | validation | warning | validation, rag, folder-sync, ui |
| gguf_filename is required: the native engine loads… | validation | error | validation, api-usage, gguf, sd-cpp |
| learning_rate must be a number, got | validation | error | training, config, hyperparameters, validation |
| STT model must be one of Studio's defaults or a Hugging… | http | error | stt, validation, model-id, huggingface |
| API key cannot be empty | http | error | api-key, validation, providers |
| dataset appears empty or unreadable | http | error | huggingface, dataset, http-422, empty-data |
| Failed to get export logs | http | error | export, logging, http-500, polling |
| features not divisible by ConvRot group | exception | error | convrot, shape-mismatch, tensor, quantization |
| Provider base URL must use http or https. | validation | error | validation, url-scheme, provider-config |
| "tool_call_id" is only valid on role="tool" messages. | validation | error | pydantic, validation, openai, chat-completions, tool-call-id, agents |
| Invalid repo_id format | http | error | hf-cache, datasets, input-validation, http-400 |
| prompts and seeds must have the same length | validation | error | diffusion, batching, seeds, validation, length-mismatch |
| ControlNet ' ' has no repo | validation | error | controlnet, catalog, config-mistake, diffusion |
| ConvRot group size must be a power of 4, got | exception | error | convrot, config-mistake, quantization, offline-build |
| dataset_streaming with evaluation requires a separate… | http | error | training, dataset-streaming, evaluation, split, http-422 |
| save_directory is required | validation | warning | validation, export, api, http-422, filesystem |
| Native path grant secret is invalid. | validation | error | configuration, base64, environment, secrets |
| activation rotation names | exception | error | convrot, checkpoint-mismatch, fqn, quantization |
| Invalid base64 image data | validation | error | diffusion, image, base64, validation, client-error |
| Provider base URL is malformed. | validation | error | validation, url-parsing, provider-config |
| a .gguf checkpoint needs model_kind 'gguf', not… | validation | error | validation, gguf, single-file, load-kinds, diffusion |
| each reference image must be at most 32 MiB (base64) | validation | error | pydantic, validation, diffusion, payload-limit, base64, reference-images |
| "name" is only valid on role="tool" messages. | validation | error | pydantic, validation, openai, chat-completions, legacy-schema |