Include token count in api.Chunk

And calculate the tokens/chunk for gemini responses, fixing the tok/s
meter for gemini models.

Further, only consider the first candidate of streamed gemini responses.

pkg/api/provider/ollama/ollama.go

@@ -182,7 +182,8 @@ func (c *OllamaClient) CreateChatCompletionStream(
 
 		if len(streamResp.Message.Content) > 0 {
 			output <- api.Chunk{
-				Content: streamResp.Message.Content,
+				Content:    streamResp.Message.Content,
+				TokenCount: 1,
 			}
 			content.WriteString(streamResp.Message.Content)
 		}