Repository Issue Activity (beta)

#86 fara 27b gguf run good

#85 does it support computer use scenairo?

#82 Hosting: llama.cpp needs `--reasoning-format none`

#81 Verify evals on Papers with Code

#79 Fara1.5 videos - do not close

#78 Stupid content filter

#75 How to train on custom tasks?

#72 Unable to reproduce Fara-7B's results on WebTailBench, even on shopping_head tasks that don't have staleness issues.

#71 Will the Fara-7B trajectories on WebTailBench be released?

#63 Release CUAVerifierBench and Universal Verifier on Hugging Face

#61 how to use api from server ?

#18 Ideas for next version

#59 Max rounds is currently hardcoded so argument doesn't apply

#58 Ollama/LM Studio hosting section could use more setup detail and compatibility notes

#57 vLLM 0.15.1 crashes on T4 GPU — CUTLASS DSL fails to detect sm_75 compute capability

#56 Trajectory logs not saved in webeval/scripts/webvoyager.py

#50 Error Provisioning Fara 7B in Microsoft Foundry

#46 Run on 16GB VRAM

#35 Feature request: Add to Microsoft Foundry Local

#49 WebVoyager eval crashes: Playwright/Chromium pthread_create: Resource temporarily unavailable

#48 No se puede instalar en la shell de fish

#39 [Config] where or what should I config in the endpoint_configs?

#42 APIConnectionError: Connection error..

#43 Problem with downloading the model

#22 How to fine-tune base Fara model on custom trajectories?