Skip to content

Add benchmark results for deepseek/deepseek-r1-distill-llama-70b#571

Merged
pei-tinybird merged 1 commit into
mainfrom
benchmark/deep-eek-deep-eek-r1-di-till-llama-70b-27129781371
Jun 8, 2026
Merged

Add benchmark results for deepseek/deepseek-r1-distill-llama-70b#571
pei-tinybird merged 1 commit into
mainfrom
benchmark/deep-eek-deep-eek-r1-di-till-llama-70b-27129781371

Conversation

@github-actions

@github-actions github-actions Bot commented Jun 8, 2026

Copy link
Copy Markdown
Contributor

This PR adds benchmark results for the deepseek/deepseek-r1-distill-llama-70b model.

Results have been pushed to Tinybird with validated=0 (pending review).
Merging this PR will validate the results, making them visible on the production dashboard.

This PR was automatically generated by the benchmark workflow.

Note: If you don't want to merge this PR, close it and the model will be added to the failed list.

/cc pei-tinybird,gnzjgo

@vercel

vercel Bot commented Jun 8, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
llm-benchmark Ready Ready Preview, Comment Jun 8, 2026 10:20am

@github-actions

github-actions Bot commented Jun 8, 2026

Copy link
Copy Markdown
Contributor Author

Automated Benchmark Review

Model: deepseek/deepseek-r1-distill-llama-70b
Review model: openai/gpt-5.4-nano
Success rate: 66.0%


  1. Quality summary: Out of 50 runs, 33 succeeded for a 66% success rate, with 48% first-attempt success, indicating a moderate level of reliability. Latency is reasonable (avg 23.64), and successful executions complete quickly (avg execution time 0.4920), suggesting performance is efficient when it works.

  2. Concerns/anomalies: The main concern is the relatively high failure count (17/50 = 34% errors), and the gap between firstAttemptRate (48%) and successRate (66%) suggests many recoveries require reruns or retries. Without error breakdowns, it’s unclear whether failures are due to model correctness, prompt/tooling issues, or runtime constraints.

  3. Recommendation: REVIEW
    Recommendation: REVIEW


This review was automatically generated. Set AUTO_MERGE=false in repo variables to disable auto-merge.

@pei-tinybird pei-tinybird merged commit 54ef9a7 into main Jun 8, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants