Spaces:
Paused
Paused
ffreemt
commited on
Commit
·
b322b9c
1
Parent(s):
dd9518b
Update README.md
Browse files
README.md
CHANGED
|
@@ -12,4 +12,12 @@ hf_oauth: true
|
|
| 12 |
hf_oauth_expiration_minutes: 480
|
| 13 |
---
|
| 14 |
|
| 15 |
-
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
hf_oauth_expiration_minutes: 480
|
| 13 |
---
|
| 14 |
|
| 15 |
+
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
| 16 |
+
|
| 17 |
+
# A Smolagents-based Agent
|
| 18 |
+
|
| 19 |
+
All free services (searxng and meta's free llama4 API)
|
| 20 |
+
|
| 21 |
+
Rather messy but should get the job done? No attempts at questioins with attched files/images/audio/video
|
| 22 |
+
|
| 23 |
+
Tried many free LLM APIs (siliconflow, pooled free-tier Gemini and Openai APIs etc.), rate limiting poses a major problem. grok-3-beta would have worked. sophnet's API also seem to work. Meta's Llama4 free APIs work quite smoothly and hopefully will score the assignment a pass 😳.
|
app.py
CHANGED
|
@@ -94,7 +94,7 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
|
|
| 94 |
"mcp-searxng"
|
| 95 |
],
|
| 96 |
"env": {
|
| 97 |
-
"SEARXNG_URL": "https://searx.
|
| 98 |
}
|
| 99 |
}
|
| 100 |
)
|
|
|
|
| 94 |
"mcp-searxng"
|
| 95 |
],
|
| 96 |
"env": {
|
| 97 |
+
"SEARXNG_URL": os.getenv("SEARXNG_URL", "https://searx.be") # https://searx.space or run and set your own
|
| 98 |
}
|
| 99 |
}
|
| 100 |
)
|