[HN Gopher] Anthropic's Haiku Beats GPT-4 Turbo in Tool Use
       ___________________________________________________________________
        
       Anthropic's Haiku Beats GPT-4 Turbo in Tool Use
        
       Author : Joschkabraun
       Score  : 36 points
       Date   : 2024-04-08 17:23 UTC (5 hours ago)
        
 (HTM) web link (docs.parea.ai)
 (TXT) w3m dump (docs.parea.ai)
        
       | yinser wrote:
       | This is great to see another player enter the tool use arena but
       | anecdotally I'm having issues getting function calls to Anthropic
       | to return proper JSON. Still feels a little fragile but hopefully
       | they continue improving.
        
         | Joschkabraun wrote:
         | Have you tried the new beta tool use API? In the experiments I
         | ran there were almost no issues parsing the function call
         | response (similar to GPT-3.5-turbo & GPT-4 turbo)
        
         | fudoshin2596 wrote:
         | Yea, new tool calling now in public beta is way better. And
         | avoids all of that XML they had before.
        
       | airocker wrote:
       | Are tools fine tuned in both or just prompt based? For a fair
       | comparison
        
         | Joschkabraun wrote:
         | All models got the same prompt fed which was essentially
         | "Question: {question}". And then the API's accept the function
         | call definition
        
           | airocker wrote:
           | But what if there is a bad model that fine tunes with only
           | tools?
        
             | Joschkabraun wrote:
             | What do you mean with being a bad model?
             | 
             | If the model is really good at tool use, then it will
             | broadly useful as it needs capabilities to generate the
             | tool definition. So, there should be some transferability.
        
         | fudoshin2596 wrote:
         | No fine tuning. Looks like he's do raw model capabilities with
         | simple prompt. repo: https://github.com/parea-ai/tool-use-
         | benchmark
        
       | mritchie712 wrote:
       | Not sure when they implemented this, but ollama now has a JSON
       | mode [0]. Not function calling, but one of the simpler ways to
       | get JSON in a local LLM. I'm using it with `knoopx/hermes-2-pro-
       | mistral:7b-q8_0` and it's worked well for me so far.
       | response = ollama.chat(model=OLLAMA_MODEL,
       | messages=[                 {                     'role':
       | 'system',                     'content': system_message,
       | },                 {                     'role': 'user',
       | 'content': user_prompt,                 },             ],
       | format='json',             options = {
       | #'temperature': 1.5, # very creative
       | 'temperature': 0.0             }             )
       | 
       | 0 -
       | https://github.com/ollama/ollama/blob/main/docs/api.md#json-...
        
         | Joschkabraun wrote:
         | Interesting. Do you have any benchmarks?
        
           | mritchie712 wrote:
           | No benchmarks, just my anecdotal experience trying to get
           | local LLM's to respond with JSON. The method above works for
           | my use case nearly 100% of the time. Other things I've tried
           | (e.g. `outlines`[0]) are really slow or don't work at all.
           | Would love to hear what others have tried!
           | 
           | 0 - https://github.com/outlines-dev/outlines
        
             | Joschkabraun wrote:
             | Ah yes. Have you tried out instructor [0] or Guidance [1]?
             | 
             | [0]: https://github.com/jxnl/instructor/
             | 
             | [1]: https://github.com/guidance-ai/guidance/tree/main
        
       ___________________________________________________________________
       (page generated 2024-04-08 23:02 UTC)