Skip to content

Commit 0f7bd30

Browse files
jeremyfowersclaude
jeremyfowers
and
claude
authored
docs(backends): drop the model catalog tables from the backend reference (#3103)
The `## Models` section duplicated `server_models.json` as ~285 lines of markdown tables (58% of the file) that nothing links to. The user-facing catalog at lemonade-server.ai/models.html fetches `server_models.json` live from GitHub raw at the release tag, so it never read these tables. Every model-catalog PR had to rebuild lemond in the `backend-docs-drift` job and regenerate the doc, or fail CI — #2495 was a PR whose entire content was three table rows added to fix that failure. Removes the section, `render_models`, the now-unused `SERVER_MODELS` constant, and the `backend-models` marker region from the template. The descriptor-backed regions (overview, support matrix, recipe options) are unchanged; those are derived from C++ and genuinely need the drift check. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
1 parent 4bf5845 commit 0f7bd30

2 files changed

Lines changed: 1 addition & 326 deletions

File tree

‎docs/dev/backends-reference.md‎

Lines changed: 0 additions & 290 deletions
Original file line numberDiff line numberDiff line change
@@ -182,296 +182,6 @@ the generator instead. Prose outside the markers is preserved. -->
182182
| `whispercpp_args` | `--whispercpp-args` | ARGS | "" | Custom arguments to pass to whisper-server |
183183
<!-- END GENERATED: backend-options -->
184184

185-
## Models
186-
187-
<!-- BEGIN GENERATED: backend-models -->
188-
#### `acestep` — ACE-Step (1 models)
189-
190-
| Model | Size (GB) | Labels |
191-
|-------|-----------|--------|
192-
| `ACE-Step-Music` | 10.5 | audio-generation |
193-
194-
#### `collection.omni` — collection.omni (5 models)
195-
196-
| Model | Size (GB) | Labels |
197-
|-------|-----------|--------|
198-
| `LMX-Omni-5.5B-Lite` | 9.3 | — |
199-
| `LMX-Omni-52B-Halo` | 44.77 | — |
200-
| `Lite Collection` | | — |
201-
| `RPG-HaloTales-V1` | 39.77 | — |
202-
| `Ultra Collection` | | — |
203-
204-
#### `kokoro` — Kokoro (1 models)
205-
206-
| Model | Size (GB) | Labels |
207-
|-------|-----------|--------|
208-
| `kokoro-v1` | 0.354 | tts |
209-
210-
#### `llamacpp` — Llama.cpp GPU (88 models)
211-
212-
| Model | Size (GB) | Labels |
213-
|-------|-----------|--------|
214-
| `Bonsai-1.7B-gguf` | 0.25 | llamacpp |
215-
| `Bonsai-4B-gguf` | 0.572 | llamacpp |
216-
| `Bonsai-8B-gguf` | 1.16 | llamacpp |
217-
| `Cogito-v2-llama-109B-MoE-GGUF` | 65.4 | vision |
218-
| `DeepSeek-Qwen3-8B-GGUF` | 5.25 | reasoning |
219-
| `Devstral-Small-2507-GGUF` | 14.3 | coding, tool-calling |
220-
| `GLM-4.5-Air-UD-Q4K-XL-GGUF` | 67.7 | reasoning |
221-
| `GLM-4.7-Flash-GGUF` | 17.5 | tool-calling |
222-
| `Gemma-3-4b-it-GGUF` | 3.34 | vision |
223-
| `Gemma-4-12B-it-GGUF` | 7.29 | tool-calling, vision, llamacpp |
224-
| `Gemma-4-12B-it-MTP-GGUF` | 7.75 | tool-calling, llamacpp, vision, mtp |
225-
| `Gemma-4-26B-A4B-it-GGUF` | 18.1 | hot, tool-calling, vision, llamacpp |
226-
| `Gemma-4-26B-A4B-it-MTP-GGUF` | 18.5 | hot, tool-calling, vision, llamacpp, mtp |
227-
| `Gemma-4-31B-it-GGUF` | 19.5 | hot, tool-calling, vision, llamacpp |
228-
| `Gemma-4-31B-it-MTP-GGUF` | 20.0 | hot, tool-calling, vision, llamacpp, mtp |
229-
| `Gemma-4-E2B-it-GGUF` | 4.09 | tool-calling, vision, llamacpp |
230-
| `Gemma-4-E4B-it-GGUF` | 5.97 | tool-calling, vision, llamacpp |
231-
| `Jan-nano-128k-GGUF` | 2.5 | — |
232-
| `Jan-v1-4B-GGUF` | 2.5 | — |
233-
| `LFM2-1.2B-GGUF` | 0.731 | — |
234-
| `LFM2-24B-A2B-GGUF` | 14.4 | — |
235-
| `LFM2-8B-A1B-GGUF` | 5.04 | — |
236-
| `LFM2.5-1.2B-Instruct-GGUF` | 0.731 | — |
237-
| `LFM2.5-8B-A1B` | 5.16 | — |
238-
| `Llama-3.2-1B-Instruct-GGUF` | 0.834 | — |
239-
| `Llama-3.2-3B-Instruct-GGUF` | 2.06 | — |
240-
| `Llama-4-Scout-17B-16E-Instruct-GGUF` | 63.2 | vision |
241-
| `MiniCPM-V-4-GGUF` | 3.15 | vision |
242-
| `MiniCPM-V-4.5-GGUF` | 6.12 | vision |
243-
| `MiniCPM-V-4.6-GGUF` | 1.64 | hot, vision |
244-
| `MiniCPM-V-4.6-Thinking-GGUF` | 1.64 | reasoning, vision |
245-
| `MiniCPM-o-2.6-Vision-GGUF` | 5.73 | vision |
246-
| `MiniCPM-o-4.5-Vision-GGUF` | 6.12 | vision |
247-
| `MiniCPM3-4B-GGUF` | 2.47 | — |
248-
| `MiniCPM4-8B-GGUF` | 4.97 | — |
249-
| `MiniCPM4.1-8B-GGUF` | 4.97 | reasoning |
250-
| `MiniCPM5-1B-GGUF` | 0.69 | hot, reasoning |
251-
| `Ministral-3-3B-Instruct-2512-GGUF` | 2.99 | vision |
252-
| `Muse-Glimmer-30B-GGUF` | 19.9 | hot, tool-calling, vision, reasoning, llamacpp, dflash |
253-
| `Nemotron-3-Nano-30B-A3B-GGUF` | 22.8 | — |
254-
| `Phi-4-mini-instruct-GGUF` | 2.49 | — |
255-
| `Playable1-GGUF` | 4.68 | coding |
256-
| `PromptBridge-0.6b-Alpha-GGUF` | 0.397 | — |
257-
| `Qwen2.5-Coder-32B-Instruct-GGUF` | 19.9 | coding |
258-
| `Qwen2.5-Omni-3B-GGUF` | 4.73 | vision, chat-transcription |
259-
| `Qwen2.5-Omni-7B-GGUF` | 7.33 | vision, chat-transcription |
260-
| `Qwen2.5-VL-3B-Instruct-GGUF` | 3.27 | vision |
261-
| `Qwen2.5-VL-7B-Instruct-GGUF` | 6.04 | vision |
262-
| `Qwen3-0.6B-GGUF` | 0.38 | reasoning |
263-
| `Qwen3-1.7B-GGUF` | 1.06 | reasoning |
264-
| `Qwen3-14B-GGUF` | 8.54 | reasoning |
265-
| `Qwen3-30B-A3B-GGUF` | 17.4 | reasoning |
266-
| `Qwen3-30B-A3B-Instruct-2507-GGUF` | 17.4 | tool-calling |
267-
| `Qwen3-4B-GGUF` | 2.38 | reasoning |
268-
| `Qwen3-4B-Instruct-2507-GGUF` | 2.5 | tool-calling |
269-
| `Qwen3-8B-GGUF` | 5.25 | reasoning |
270-
| `Qwen3-Coder-30B-A3B-Instruct-GGUF` | 18.6 | coding, tool-calling, hot |
271-
| `Qwen3-Coder-Next-GGUF` | 48.0 | coding, tool-calling, hot |
272-
| `Qwen3-Embedding-0.6B-GGUF` | 0.64 | embeddings |
273-
| `Qwen3-Embedding-4B-GGUF` | 4.28 | embeddings |
274-
| `Qwen3-Embedding-8B-GGUF` | 8.05 | embeddings |
275-
| `Qwen3-Next-80B-A3B-Instruct-GGUF` | 46.1 | tool-calling |
276-
| `Qwen3-VL-4B-Instruct-GGUF` | 3.33 | vision |
277-
| `Qwen3-VL-8B-Instruct-GGUF` | 6.19 | vision |
278-
| `Qwen3.5-0.8B-GGUF` | 0.764 | vision, tool-calling |
279-
| `Qwen3.5-122B-A10B-GGUF` | 77.9 | vision, tool-calling |
280-
| `Qwen3.5-122B-A10B-MTP-GGUF` | 79.6 | vision, tool-calling, mtp |
281-
| `Qwen3.5-27B-GGUF` | 18.5 | vision, tool-calling |
282-
| `Qwen3.5-2B-GGUF` | 2.01 | vision, tool-calling |
283-
| `Qwen3.5-35B-A3B-GGUF` | 23.1 | vision, tool-calling |
284-
| `Qwen3.5-4B-GGUF` | 3.58 | vision, tool-calling, hot |
285-
| `Qwen3.5-4B-MTP-GGUF` | 3.66 | vision, tool-calling, mtp |
286-
| `Qwen3.5-9B-GGUF` | 6.88 | vision, tool-calling |
287-
| `Qwen3.6-27B-GGUF` | 18.5 | vision, tool-calling |
288-
| `Qwen3.6-27B-MTP-GGUF` | 18.8 | vision, tool-calling, mtp, hot |
289-
| `Qwen3.6-35B-A3B-GGUF` | 23.3 | vision, tool-calling, hot |
290-
| `Qwen3.6-35B-A3B-MTP-GGUF` | 23.8 | vision, tool-calling, mtp |
291-
| `SmolLM3-3B-GGUF` | 1.94 | — |
292-
| `Tiny-Test-Model-GGUF` | 0.18 | — |
293-
| `bge-reranker-v2-m3-GGUF` | 0.636 | reranking |
294-
| `gpt-oss-120b-GGUF` | 62.8 | reasoning, tool-calling |
295-
| `gpt-oss-120b-mxfp-GGUF` | 63.4 | hot, reasoning, tool-calling |
296-
| `gpt-oss-20b-GGUF` | 11.6 | reasoning, tool-calling |
297-
| `gpt-oss-20b-mxfp4-GGUF` | 12.1 | hot, reasoning, tool-calling |
298-
| `granite-4.0-h-tiny-GGUF` | 4.25 | tool-calling |
299-
| `jina-reranker-v1-tiny-en-GGUF` | 0.0367 | reranking |
300-
| `nomic-embed-text-v1-GGUF` | 0.0781 | embeddings |
301-
| `nomic-embed-text-v2-moe-GGUF` | 0.51 | embeddings |
302-
303-
#### `moonshine` — Moonshine (3 models)
304-
305-
| Model | Size (GB) | Labels |
306-
|-------|-----------|--------|
307-
| `Moonshine-Medium-Streaming` | 1.08 | transcription, realtime-transcription, hot |
308-
| `Moonshine-Small-Streaming` | 0.431 | transcription, realtime-transcription |
309-
| `Moonshine-Tiny-Streaming` | 0.202 | transcription, realtime-transcription |
310-
311-
#### `onnxruntime` — ONNX Runtime (2 models)
312-
313-
| Model | Size (GB) | Labels |
314-
|-------|-----------|--------|
315-
| `Bert-Phishing-ONNX` | 1.34 | classification |
316-
| `Phishing-Email-Detection-ONNX` | 0.27 | classification |
317-
318-
#### `openmoss` — OpenMOSS TTS (2 models)
319-
320-
| Model | Size (GB) | Labels |
321-
|-------|-----------|--------|
322-
| `MOSS-VoiceGen` | 7.3 | tts, voice-design |
323-
| `OpenMOSS-TTS` | 12.5 | tts |
324-
325-
#### `ryzenai-llm` — Ryzen AI LLM (79 models)
326-
327-
| Model | Size (GB) | Labels |
328-
|-------|-----------|--------|
329-
| `AMD-OLMo-1B-SFT-DPO-Hybrid` | 1.48 | — |
330-
| `CodeLlama-7b-Instruct-hf-Hybrid` | 7.24 | coding |
331-
| `CodeLlama-7b-Instruct-hf-NPU` | 7.54 | coding |
332-
| `DeepSeek-R1-Distill-Llama-8B-CPU` | 6.2 | reasoning |
333-
| `DeepSeek-R1-Distill-Llama-8B-Hybrid` | 9.09 | reasoning |
334-
| `DeepSeek-R1-Distill-Llama-8B-NPU` | 9.3 | reasoning |
335-
| `DeepSeek-R1-Distill-Qwen-1.5B-Hybrid` | 2.19 | reasoning |
336-
| `DeepSeek-R1-Distill-Qwen-1.5B-NPU` | 2.3 | reasoning |
337-
| `DeepSeek-R1-Distill-Qwen-7B-CPU` | 6.2 | reasoning |
338-
| `DeepSeek-R1-Distill-Qwen-7B-Hybrid` | 8.67 | reasoning |
339-
| `DeepSeek-R1-Distill-Qwen-7B-NPU` | 8.87 | reasoning |
340-
| `Gemma-3-4b-it-mm-NPU` | 6.68 | vision |
341-
| `Llama-2-7b-chat-hf-Hybrid` | 7.31 | — |
342-
| `Llama-2-7b-chat-hf-NPU` | 7.47 | — |
343-
| `Llama-2-7b-hf-Hybrid` | 7.31 | — |
344-
| `Llama-2-7b-hf-NPU` | 7.47 | — |
345-
| `Llama-3.1-8B-Hybrid` | 9.09 | — |
346-
| `Llama-3.1-8B-NPU` | 9.3 | — |
347-
| `Llama-3.2-1B-Hybrid` | 1.89 | — |
348-
| `Llama-3.2-1B-Instruct-CPU` | 1.76 | — |
349-
| `Llama-3.2-1B-Instruct-Hybrid` | 1.89 | — |
350-
| `Llama-3.2-1B-Instruct-NPU` | 1.96 | — |
351-
| `Llama-3.2-1B-NPU` | 1.96 | — |
352-
| `Llama-3.2-3B-Hybrid` | 4.28 | — |
353-
| `Llama-3.2-3B-Instruct-CPU` | 3.38 | — |
354-
| `Llama-3.2-3B-Instruct-Hybrid` | 4.28 | — |
355-
| `Meta-Llama-3-8B-Hybrid` | 9.06 | — |
356-
| `Meta-Llama-3-8B-NPU` | 9.23 | — |
357-
| `Meta-Llama-3.1-8B-Instruct-Hybrid` | 9.09 | — |
358-
| `Meta-Llama-3.1-8B-Instruct-NPU` | 9.3 | — |
359-
| `Mistral-7B-Instruct-v0.1-Hybrid` | 7.34 | — |
360-
| `Mistral-7B-Instruct-v0.1-NPU` | 8.01 | — |
361-
| `Mistral-7B-Instruct-v0.2-Hybrid` | 7.34 | — |
362-
| `Mistral-7B-Instruct-v0.2-NPU` | 8.01 | — |
363-
| `Mistral-7B-Instruct-v0.3-Hybrid` | 7.35 | — |
364-
| `Mistral-7B-Instruct-v0.3-NPU` | 8.09 | — |
365-
| `Mistral-7B-v0.3-Hybrid` | 7.35 | — |
366-
| `Mistral-7B-v0.3-NPU` | 8.09 | — |
367-
| `Phi-3-Mini-Instruct-CPU` | 2.39 | — |
368-
| `Phi-3-mini-128k-instruct-Hybrid` | 4.21 | — |
369-
| `Phi-3-mini-128k-instruct-NPU` | 4.35 | — |
370-
| `Phi-3-mini-4k-instruct-Hybrid` | 4.19 | — |
371-
| `Phi-3-mini-4k-instruct-NPU` | 4.3 | — |
372-
| `Phi-3.5-mini-instruct-Hybrid` | 4.21 | — |
373-
| `Phi-3.5-mini-instruct-NPU` | 4.35 | — |
374-
| `Phi-4-mini-instruct-Hybrid` | 5.47 | — |
375-
| `Phi-4-mini-instruct-NPU` | 5.59 | — |
376-
| `Phi-4-mini-reasoning-Hybrid` | 5.47 | reasoning |
377-
| `Qwen-1.5-7B-Chat-CPU` | 6.32 | — |
378-
| `Qwen-2.5-1.5B-Instruct-Hybrid` | 2.17 | — |
379-
| `Qwen-2.5-1.5B-Instruct-NPU` | 2.25 | — |
380-
| `Qwen1.5-7B-Chat-Hybrid` | 8.83 | — |
381-
| `Qwen1.5-7B-Chat-NPU` | 9.02 | — |
382-
| `Qwen2-1.5B-Hybrid` | 2.19 | — |
383-
| `Qwen2-1.5B-NPU` | 2.3 | — |
384-
| `Qwen2-7B-Hybrid` | 8.68 | — |
385-
| `Qwen2-7B-NPU` | 8.88 | — |
386-
| `Qwen2.5-0.5B-Instruct-CPU` | 0.834 | — |
387-
| `Qwen2.5-0.5B-Instruct-Hybrid` | 0.828 | — |
388-
| `Qwen2.5-14B-instruct-Hybrid` | 16.5 | — |
389-
| `Qwen2.5-3B-Instruct-Hybrid` | 3.97 | — |
390-
| `Qwen2.5-3B-Instruct-NPU` | 4.1 | — |
391-
| `Qwen2.5-7B-Instruct-Hybrid` | 8.65 | — |
392-
| `Qwen2.5-7B-Instruct-NPU` | 8.83 | — |
393-
| `Qwen2.5-Coder-0.5B-Instruct-Hybrid` | 0.828 | coding |
394-
| `Qwen2.5-Coder-1.5B-Instruct-Hybrid` | 2.17 | coding |
395-
| `Qwen2.5-Coder-1.5B-Instruct-NPU` | 2.25 | coding |
396-
| `Qwen2.5-Coder-7B-Instruct-Hybrid` | 8.65 | coding |
397-
| `Qwen2.5-Coder-7B-Instruct-NPU` | 8.83 | coding |
398-
| `Qwen3-1.7B-Hybrid` | 2.55 | reasoning |
399-
| `Qwen3-14B-Hybrid` | 16.5 | reasoning |
400-
| `Qwen3-4B-Hybrid` | 5.17 | reasoning |
401-
| `Qwen3-8B-Hybrid` | 9.42 | reasoning |
402-
| `SmolLM-135M-Instruct-Hybrid` | 0.232 | — |
403-
| `SmolLM2-135M-Instruct-Hybrid` | 0.233 | — |
404-
| `chatglm3-6b-Hybrid` | 6.9 | — |
405-
| `chatglm3-6b-NPU` | 7.04 | — |
406-
| `gemma-2-2b-Hybrid` | 4.04 | — |
407-
| `gpt-oss-20b-NPU` | 13.4 | — |
408-
409-
#### `sd-cpp` — StableDiffusion.cpp (12 models)
410-
411-
| Model | Size (GB) | Labels |
412-
|-------|-----------|--------|
413-
| `Flux-2-Klein-4B` | 16.1 | image, edit |
414-
| `Flux-2-Klein-9B-GGUF` | 19.0 | image, edit |
415-
| `Qwen-Image-2512-GGUF` | 19.4 | image |
416-
| `Qwen-Image-GGUF` | 18.2 | image |
417-
| `RealESRGAN-x4plus` | 0.064 | upscaling, image |
418-
| `RealESRGAN-x4plus-anime` | 0.017 | upscaling, image |
419-
| `SD-1.5` | 7.7 | image |
420-
| `SD-Turbo` | 5.21 | image |
421-
| `SD-Turbo-GGUF` | 2.02 | image |
422-
| `SDXL-Base-1.0` | 6.94 | image |
423-
| `SDXL-Turbo` | 6.94 | image |
424-
| `Z-Image-Turbo` | 20.7 | image |
425-
426-
#### `thenoise` — TheNoise ROCm (experimental) (4 models)
427-
428-
| Model | Size (GB) | Labels |
429-
|-------|-----------|--------|
430-
| `Anima-Aesthetic` | 5.4 | image |
431-
| `Anima-Base` | 5.4 | image |
432-
| `Anima-Turbo` | 5.4 | image |
433-
| `Krea-2-Turbo` | 35 | image |
434-
435-
#### `thinksound` — ThinkSound (1 models)
436-
437-
| Model | Size (GB) | Labels |
438-
|-------|-----------|--------|
439-
| `ThinkSound-SFX` | 6.4 | audio-generation |
440-
441-
#### `trellis` — TRELLIS.2 (1 models)
442-
443-
| Model | Size (GB) | Labels |
444-
|-------|-----------|--------|
445-
| `TRELLIS-3D` | 15.4 | 3d |
446-
447-
#### `vllm` — vLLM ROCm (experimental) (11 models)
448-
449-
| Model | Size (GB) | Labels |
450-
|-------|-----------|--------|
451-
| `GLM-4.7-Flash-FP16-vLLM` | 62.47 | reasoning, tool-calling |
452-
| `Qwen3.5-0.8B-FP16-vLLM` | 1.77 | reasoning |
453-
| `Qwen3.5-2B-FP16-vLLM` | 4.57 | reasoning, tool-calling |
454-
| `Qwen3.5-4B-FP16-vLLM` | 9.34 | reasoning, hot, tool-calling |
455-
| `Qwen3.5-9B-FP16-vLLM` | 19.3 | reasoning, tool-calling |
456-
| `Qwen3.6-27B-FP16-vLLM` | 55.59 | reasoning, tool-calling, vision |
457-
| `Qwen3.6-27B-FP8-vLLM-highconc` | 27.8 | reasoning, tool-calling, vision, mtp |
458-
| `Qwen3.6-27B-FP8-vLLM-lowconc` | 27.8 | reasoning, tool-calling, vision, mtp |
459-
| `Qwen3.6-35B-A3B-FP16-vLLM` | 71.93 | reasoning, tool-calling, vision |
460-
| `Qwen3.6-35B-A3B-FP8-vLLM-highconc` | 36.0 | reasoning, tool-calling, vision, mtp |
461-
| `Qwen3.6-35B-A3B-FP8-vLLM-lowconc` | 36.0 | reasoning, tool-calling, vision, mtp |
462-
463-
#### `whispercpp` — Whisper.cpp (6 models)
464-
465-
| Model | Size (GB) | Labels |
466-
|-------|-----------|--------|
467-
| `Whisper-Base` | 0.148 | transcription, realtime-transcription |
468-
| `Whisper-Large-v3` | 3.1 | transcription, realtime-transcription |
469-
| `Whisper-Large-v3-Turbo` | 1.62 | transcription, realtime-transcription, hot |
470-
| `Whisper-Medium` | 1.53 | transcription, realtime-transcription |
471-
| `Whisper-Small` | 0.488 | transcription, realtime-transcription |
472-
| `Whisper-Tiny` | 0.075 | transcription, realtime-transcription |
473-
<!-- END GENERATED: backend-models -->
474-
475185
## Implementation notes
476186

477187
### ACE-Step (`acestep`)

‎docs/tools/gen_backend_boilerplate.py‎

Lines changed: 1 addition & 36 deletions
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@
77
hand-maintained:
88
99
* Marker-delimited regions of the backend reference docs, from
10-
``/system-info`` ``recipes`` + ``server_models.json``.
10+
``/system-info`` ``recipes``.
1111
* The whole of ``src/cpp/resources/defaults.json``, mirrored verbatim from
1212
``/internal/config/defaults`` (its per-recipe blocks come from each
1313
descriptor's ``config_defaults()``).
@@ -38,7 +38,6 @@
3838
from pathlib import Path
3939

4040
REPO_ROOT = Path(__file__).resolve().parents[2]
41-
SERVER_MODELS = REPO_ROOT / "src" / "cpp" / "resources" / "server_models.json"
4241
TARGET_DOC = REPO_ROOT / "docs" / "dev" / "backends-reference.md"
4342

4443

@@ -430,34 +429,6 @@ def render_options(recipes: dict) -> str:
430429
return "\n".join(blocks).rstrip()
431430

432431

433-
def render_models(recipes: dict) -> str:
434-
models = json.loads(SERVER_MODELS.read_text())
435-
by_recipe: dict[str, list] = {}
436-
for name, data in models.items():
437-
if not isinstance(data, dict):
438-
continue
439-
by_recipe.setdefault(data.get("recipe", "(unspecified)"), []).append(
440-
(name, data)
441-
)
442-
blocks = []
443-
for recipe in sorted(by_recipe):
444-
entries = sorted(by_recipe[recipe])
445-
display = recipes.get(recipe, {}).get("display_name", recipe)
446-
blocks.append(f"#### `{recipe}` — {display} ({len(entries)} models)\n")
447-
blocks.append("| Model | Size (GB) | Labels |")
448-
blocks.append("|-------|-----------|--------|")
449-
for name, data in entries:
450-
blocks.append(
451-
"| `{n}` | {s} | {l} |".format(
452-
n=md_escape(name),
453-
s=data.get("size", ""),
454-
l=md_escape(", ".join(data.get("labels", []))) or "—",
455-
)
456-
)
457-
blocks.append("")
458-
return "\n".join(blocks).rstrip()
459-
460-
461432
DEFAULT_TEMPLATE = """# Backend reference
462433
463434
<!-- This file is generated by docs/tools/gen_backend_boilerplate.py from the C++ backend
@@ -478,11 +449,6 @@ def render_models(recipes: dict) -> str:
478449
479450
<!-- BEGIN GENERATED: backend-options -->
480451
<!-- END GENERATED: backend-options -->
481-
482-
## Models
483-
484-
<!-- BEGIN GENERATED: backend-models -->
485-
<!-- END GENERATED: backend-models -->
486452
"""
487453

488454

@@ -551,7 +517,6 @@ def main() -> int:
551517
"backends-overview": render_overview(recipes),
552518
"backends-matrix": render_support_matrix(recipes),
553519
"backend-options": render_options(recipes),
554-
"backend-models": render_models(recipes),
555520
},
556521
"template": DEFAULT_TEMPLATE,
557522
},

0 commit comments

Comments
 (0)