Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -52,6 +52,10 @@ The same pipeline is available three ways:
|
|
| 52 |
2. **MCP server**: add `https://opceanai-pasita.hf.space/gradio_api/mcp/` to any MCP client.
|
| 53 |
3. **Gradio API**: the `convert` endpoint, documented in the `?view=api` drawer.
|
| 54 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
Errors use the OpenAI envelope (`{"error": {"message", "type", "param", "code"}}`); ZeroGPU
|
| 56 |
quota exhaustion answers with HTTP 429.
|
| 57 |
|
|
|
|
| 52 |
2. **MCP server**: add `https://opceanai-pasita.hf.space/gradio_api/mcp/` to any MCP client.
|
| 53 |
3. **Gradio API**: the `convert` endpoint, documented in the `?view=api` drawer.
|
| 54 |
|
| 55 |
+
Sampling temperature is fixed at 0.3 because that is the measured recipe; both REST
|
| 56 |
+
endpoints accept an optional `max_tokens` cap. The Gradio v2 call endpoint also accepts
|
| 57 |
+
the legacy `{"data": [...]}` body shape for compatibility.
|
| 58 |
+
|
| 59 |
Errors use the OpenAI envelope (`{"error": {"message", "type", "param", "code"}}`); ZeroGPU
|
| 60 |
quota exhaustion answers with HTTP 429.
|
| 61 |
|