Hello there could you share the prompt you tried?

The model by default will work as a translator unless it detects an explicit instruction.

from transformers import AutoModelForCausalLM, AutoTokenizer

Load model and tokenizer

model = AutoModelForCausalLM.from_pretrained("CraneAILabs/ganda-gemma-1b")
tokenizer = AutoTokenizer.from_pretrained("CraneAILabs/ganda-gemma-1b")

Translate to Luganda

prompt = "Translate to Luganda: Hello, how are you today?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=100, temperature=0.3)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

I got this response

Translate to Luganda: Hello, how are you today?

@sulaimank this seems a little strange, I can't recreate it on my end. Can you rerun the command a couple of times and let us know if its consistent?

I actually ran the quick start code in the model card. I was testing it out. Though it seems to return the prompt itself on my side.

Here is the quick start code I ran from the model card

from transformers import AutoModelForCausalLM, AutoTokenizer

Load model and tokenizer

model = AutoModelForCausalLM.from_pretrained("CraneAILabs/ganda-gemma-1b")
tokenizer = AutoTokenizer.from_pretrained("CraneAILabs/ganda-gemma-1b")

Translate to Luganda

prompt = "Translate to Luganda: Hello, how are you today?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=100, temperature=0.3)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

@sulaimank we're going to advise you to try the pipeline example for now as we figure this out.
It seems to be inconsistently happening across different machines.