No official torrent yet — requesting it tells the flock what to host next. You cannot seed a catalog-only page.
Cataloged from BananaMind/BananaMind-2-assistant2user
All models by BananaMind →[Image blocked: banner] BananaMind 2 Assistant2User is a 139M parameter language model trained to reconstruct a likely user message from an assistant response.
The model starts from BananaMind/BananaMind-2-Pro and is fully fine-tuned on reversed conversations from HuggingFaceTB/smol-smoltalk.
Normal conversation direction:
Training direction:
The model receives the assistant response as context and predicts the user message that likely caused that response.
Input:
Output:
The original Smol-SmolTalk conversations contain messages in the usual structure:
Every adjacent user and assistant pair was converted into an independent reversed training example.
For example:
becomes:
Loss is applied only to the reconstructed user message. The assistant response and reconstruction prefix are masked from the training loss.
The model was evaluated by taking conversations from the Smol-SmolTalk test split, hiding the original user message, giving the model only the assistant response, and comparing the generated reconstruction against the original user message.
The results below are from the first 2500 extracted user-assistant pairs from the test split.
Token Top-1 Accuracy measures how close the generated token sequence is to the original user message using token-level edit distance.
A score of 100% means the generated user message is token-for-token identical after tokenization.
This metric is strict. Semantically equivalent reconstructions can receive a much lower score when they use different wording.
For example:
These messages express essentially the same request while differing substantially at the token level.
Assistant response:
Reconstructed user message:
Assistant response:
Reconstructed user message:
Assistant response:
Reconstructed user message:
Assistant response:
Reconstructed user message:
Assistant response:
Reconstructed user message:
Assistant response:
Reconstructed user message:
This example was tested on an assistant response from a real ChatGPT conversation rather than Smol-SmolTalk.
Example output:
Assistant-to-user reconstruction is inherently underdetermined.
An assistant response such as:
could reasonably come from many prompts:
The model therefore learns to generate a plausible originating user message based on the assistant response and the patterns present in its training data.
Short responses such as:
provide very little information about the original question. Reconstructions from highly ambiguous responses can contain invented context.
Longer and more informative assistant responses generally provide much stronger signals for reconstruction.
The model has approximately 139M parameters, so generation quality varies substantially across inputs.
Observed failure modes include:
The model performs especially well on assistant messages where the likely originating request is strongly implied by the answer.
User message↓Assistant responseAssistant response↓Reconstructed user messageAssistant answer:The capital of Japan is Tokyo.
Reconstruct the user message:What is the capital of Japan?user → assistantuser:How do I reverse a Python list?
assistant:You can reverse a Python list with list.reverse() or slicing with [::-1].Assistant answer:You can reverse a Python list with list.reverse() or slicing with [::-1].
Reconstruct the user message:How do I reverse a Python list?accuracy = 1 - token_edit_distance / max(reference_tokens, generated_tokens)Original:for what is the ml-intern-explorers org made
Reconstruction:What is the purpose of the ml-intern-explorers Hugging Face org?The capital of Japan is Tokyo.What is the capital of Japan?No, 17 is not an even number.Is 17 an even number?To reverse a Python list, you can use list.reverse() or slicing with [::-1].Write a Python program that reverses a list of integers.You can install PyTorch with pip using pip install torch.I am trying to install PyTorch on my machine. I have a Python script that I want to use to run PyTorch. I have tried to install PyTorch using pip, but it doesn't seem to work. Can you help me install PyTorch with pip?A binary search runs in O(log n) time on a sorted array.What is the time complexity of a binary search?The `ml-intern-explorers` Hugging Face org appears to have been created mainly as a shared community workspace for people/AI agents experimenting with ML...
Its clearest documented purpose is multi-agent ML collaboration...What is the purpose of the ml-intern-explorers Hugging Face org?import torchfrom transformers import AutoTokenizer, AutoModelForCausalLM
model_id = "BananaMind/BananaMind-2-assistant2user"
tokenizer = AutoTokenizer.from_pretrained( model_id, trust_remote_code=True,)
model = AutoModelForCausalLM.from_pretrained( model_id, trust_remote_code=True, dtype=torch.bfloat16,).cuda()
model.eval()
assistant_message = """The capital of Japan is Tokyo."""
prompt = ( "Assistant answer:\n" + assistant_message.strip() + "\n\n" + "Reconstruct the user message:\n")
inputs = tokenizer( prompt, return_tensors="pt", add_special_tokens=False,).to("cuda")
with torch.inference_mode(): output = model.generate( **inputs, max_new_tokens=256, do_sample=False, eos_token_id=tokenizer.eos_token_id, pad_token_id=tokenizer.eos_token_id, )
generated = output[0, inputs["input_ids"].shape[1]:]
print( tokenizer.decode( generated, skip_special_tokens=True, ))What is the capital of Japan?The color of the sky is blue.What color is the sky?What is the color of the sky?Is the sky blue?What color does the sky normally appear?Yes.