The generate POST succeeds and returns a job ID, but retrieving results always fails with HTTP 400.

Version: 0.12.1

The generate POST succeeds and returns a job ID, but retrieving
results always fails with HTTP 400.

Looking at embeddings/api.py:

• POST goes to: /api/v1/embeddings/models/{model_id}/generate ✓
• GET goes to: /api/v1/embeddings/{job_id}/generate ✗

The GET path appears to be missing the /models/{model_id}/
prefix compared to the POST path. This may be a URL mismatch.

Failing job ID: f83e723e-6f03-4c79-a46a-cc752b6a30e4
Server response: {"detail":"error retrieving results","jid":"..."}

Minimal reproduction:
poet = session.embedding.get_model("poet")
gen_job = poet.generate(prompt=prompt_id_string, num_samples=10)
gen_job.wait() # raises HTTPError 400

从 embeddings/api.py 开始，跟踪 generate job 的 POST URL 和结果获取 URL。使用文档中的最小示例，通过 poet.generate(...)，然后使用 gen_job.wait() 重现该问题。当结果获取使用正确的模型范围 endpoint 且不再因 HTTP 400 失败时，即表示完成。