Spaces:

cstr
/

LLMProviders

Paused

App Files Files Community

LLMProviders / scripts

Commit History

Mar 22

19:50
fix: refine OCR label, remove hallucinated HF IDs, and optimize gated repo validation

CrispStrobe

ffc4314
19:24
feat: simplify OCR display and optimize HF repo validation with caching

CrispStrobe

704a297
19:18
feat: verbose HF validation, OCR benchmarks, and corrected model mappings

CrispStrobe

7cc1131
18:52
fix: remove hallucinated HF IDs and improve gated repo validation

CrispStrobe

af78198
18:31
fix: resolve white screen by adding try-catch to benchmark sorting and rendering

CrispStrobe

5e7afae
17:32
feat: extract MTEB benchmarks from Hugging Face README metadata

CrispStrobe

5afa393
17:30
feat: enhance MTEB enrichment with cross-revision aggregation and manual fallbacks for latest models

CrispStrobe

062b038
16:27
feat: integrate MTEB benchmarks from official GitHub results repo

CrispStrobe

b090cc8
16:07
feat: track and save the source of each benchmark datapoint

CrispStrobe

4bcf2a4
14:41
feat: add HF IDs and sizes for Voxtral and new Mistral models

CrispStrobe

131ab2f
14:38
fix: propagate model capabilities globally and add missing icons for Whisper/Voxtral

CrispStrobe

6f9105d
14:30
feat: comprehensive audio/multimodal pricing support and improved model metadata enrichment (Whisper, Gemini 3.1, Voxtral)

CrispStrobe

ec9cfc9
13:35
fix: resolve HTTP 408 timeouts and finalize technical metadata enrichment for MiniMax-M1 and others

CrispStrobe

e16bbaf
13:33
feat: resolve MiniMax-M1 (366B), Mistral Small variants, and Qwen Plus via deep config.json parsing and global technical sync

CrispStrobe

c5e4684
13:02
fix: aggressively parse raw config.json to estimate parameters for models with minified HF metadata

CrispStrobe

e2e9c3d
12:42
feat: achieve 100% public model metadata coverage; improved vLLM-style param estimation; fixed phi-4 and mistral variants

CrispStrobe

54c2fe8
11:47
fix: resolve minified HF metadata by fetching raw config.json; improved BERT-style param estimation

CrispStrobe

128b3b8
11:10
feat: achieve 100% metadata coverage with automated proprietary model detection and expanded mappings

CrispStrobe

901bb12
11:02
feat: implement technical repo-id based metadata enrichment and grouping

CrispStrobe

f0da319
10:55
feat: implement persistent HF private status and aggressive manual mappings for Qwen, DeepSeek, and Grok

CrispStrobe

eed68bb
10:43
feat: complete full set of Artificial Analysis metrics

CrispStrobe

ac03708
10:43
feat: integrate Ollama Registry, vLLM-style param estimation, and enhanced model info tooltips

CrispStrobe

db007de
10:28
feat: integrate Artificial Analysis API and correct Mistral family sizes

CrispStrobe

cf1221a
10:11
fix: refine FLUX family sizes to 32B for FLUX.2 flagships based on official specs

CrispStrobe

388998e
10:06
fix: correct FLUX.2 family sizes to 32B and refine hf_id grouping logic

CrispStrobe

59b4e0e
09:52
feat: prioritize hf_id for model grouping and enrich FLUX family metadata

CrispStrobe

2a1a498
09:26
feat: implement private model caching for HF Hub and add manual mappings for Qwen/IONOS models

CrispStrobe

03b99dd
09:18
fix: add HF lookup circuit breaker, detailed error logging, and improved size regex

CrispStrobe

88466b5
09:14
fix: optimize HF Hub lookups with smart merging, limited retries, and detailed logging

CrispStrobe

bffabba
08:57
feat: add interactive metadata info icon with direct Hugging Face links

CrispStrobe

225d830
08:54
feat: improve size precision for small models and add manual HF mapping for Infomaniak

CrispStrobe

02e6e85
08:46
fix: prevent eu-endpoint leakage to other providers and optimize HF Hub retries

CrispStrobe

3a30924
08:17
feat: include eu-endpoint tagging in OpenRouter data

CrispStrobe

e9acff4
08:07
feat: improve size detection via hf_id inheritance, add embedder icon, and restore comments

CrispStrobe

c3fc087
07:59
feat: improve size detection (decimals, description regex) and embedder identification

CrispStrobe

d756c8e
05:33
fix: implement robust fetching with retries and exponential backoff for benchmarks and providers

CrispStrobe

d135f12

Mar 2

06:03
add README, AGPL-3 license, updated provider data

CrispStrobe

a1c1ca9
05:20
add bfl, etc

CrispStrobe

ee20373
04:58
init

CrispStrobe

ffba252

Commit History

Mar 22

fix: refine OCR label, remove hallucinated HF IDs, and optimize gated repo validation

feat: simplify OCR display and optimize HF repo validation with caching

feat: verbose HF validation, OCR benchmarks, and corrected model mappings

fix: remove hallucinated HF IDs and improve gated repo validation

fix: resolve white screen by adding try-catch to benchmark sorting and rendering

feat: extract MTEB benchmarks from Hugging Face README metadata

feat: enhance MTEB enrichment with cross-revision aggregation and manual fallbacks for latest models

feat: integrate MTEB benchmarks from official GitHub results repo

feat: track and save the source of each benchmark datapoint

feat: add HF IDs and sizes for Voxtral and new Mistral models

fix: propagate model capabilities globally and add missing icons for Whisper/Voxtral

feat: comprehensive audio/multimodal pricing support and improved model metadata enrichment (Whisper, Gemini 3.1, Voxtral)

fix: resolve HTTP 408 timeouts and finalize technical metadata enrichment for MiniMax-M1 and others

feat: resolve MiniMax-M1 (366B), Mistral Small variants, and Qwen Plus via deep config.json parsing and global technical sync

fix: aggressively parse raw config.json to estimate parameters for models with minified HF metadata

feat: achieve 100% public model metadata coverage; improved vLLM-style param estimation; fixed phi-4 and mistral variants

fix: resolve minified HF metadata by fetching raw config.json; improved BERT-style param estimation

feat: achieve 100% metadata coverage with automated proprietary model detection and expanded mappings

feat: implement technical repo-id based metadata enrichment and grouping

feat: implement persistent HF private status and aggressive manual mappings for Qwen, DeepSeek, and Grok

feat: complete full set of Artificial Analysis metrics

feat: integrate Ollama Registry, vLLM-style param estimation, and enhanced model info tooltips

feat: integrate Artificial Analysis API and correct Mistral family sizes

fix: refine FLUX family sizes to 32B for FLUX.2 flagships based on official specs

fix: correct FLUX.2 family sizes to 32B and refine hf_id grouping logic

feat: prioritize hf_id for model grouping and enrich FLUX family metadata

feat: implement private model caching for HF Hub and add manual mappings for Qwen/IONOS models

fix: add HF lookup circuit breaker, detailed error logging, and improved size regex

fix: optimize HF Hub lookups with smart merging, limited retries, and detailed logging

feat: add interactive metadata info icon with direct Hugging Face links

feat: improve size precision for small models and add manual HF mapping for Infomaniak

fix: prevent eu-endpoint leakage to other providers and optimize HF Hub retries

feat: include eu-endpoint tagging in OpenRouter data

feat: improve size detection via hf_id inheritance, add embedder icon, and restore comments

feat: improve size detection (decimals, description regex) and embedder identification

fix: implement robust fetching with retries and exponential backoff for benchmarks and providers

Mar 2

add README, AGPL-3 license, updated provider data

add bfl, etc

init