biRddlmZGddeZGddeZGddeZy) )Optimizer2Statec:eZdZ dfd ZxZS)LAMBc<t|d|||||| | | | | d y)a Base LAMB optimizer. Arguments: params (`torch.tensor`): The input parameters to optimize. lr (`float`, defaults to 1e-3): The learning rate. bias_correction (`bool`, defaults to `True`): Whether to apply bias correction to the first and second-order moments. betas (`tuple(float, float)`, defaults to (0.9, 0.999)): The beta values are the decay rates of the first and second-order moment of the optimizer. eps (`float`, defaults to 1e-8): The epsilon value prevents division by zero in the optimizer. weight_decay (`float`, defaults to 1e-2): The weight decay value for the optimizer. amsgrad (`bool`, defaults to `False`): Whether to use the [AMSGrad](https://hf.co/papers/1904.09237) variant of Adam that uses the maximum of past squared gradients instead. adam_w_mode (`bool`, defaults to `True`): Whether to use the AdamW variant. optim_bits (`int`, defaults to 32): The number of bits of the optimizer state. args (`object`, defaults to `None`): An object with additional arguments. min_8bit_size (`int`, defaults to 4096): The minimum number of elements of the parameter tensors for 8-bit optimization. percentile_clipping (`int`, defaults to 100): Adapts clipping threshold automatically by tracking the last 100 gradient norms and clipping the gradient at a certain percentile to improve stability. block_wise (`bool`, defaults to `True`): Whether to independently quantize each block of tensors to reduce outlier effects and improve stability. max_unorm (`float`, defaults to 1.0): The maximum gradient norm. lamb? max_unormNsuper__init__)selfparamslrbias_correctionbetaseps weight_decayamsgrad adam_w_mode optim_bitsargs min_8bit_sizepercentile_clipping block_wiser __class__s R/home/cdr/jupyterlab/.venv/lib/python3.12/site-packages/bitsandbytes/optim/lamb.pyr z LAMB.__init__ s<d              ) MbP?Tg?g+?:0yE>rFT NdFr__name__ __module__ __qualname__r __classcell__rs@rrrs6   ? ? rrc8eZdZ dfd ZxZS)LAMB8bitc<t|d|||||d| | | | d y)a( 8-bit LAMB optimizer. Arguments: params (`torch.tensor`): The input parameters to optimize. lr (`float`, defaults to 1e-3): The learning rate. bias_correction (`bool`, defaults to `True`): Whether to apply bias correction to the first and second-order moments. betas (`tuple(float, float)`, defaults to (0.9, 0.999)): The beta values are the decay rates of the first and second-order moment of the optimizer. eps (`float`, defaults to 1e-8): The epsilon value prevents division by zero in the optimizer. weight_decay (`float`, defaults to 1e-2): The weight decay value for the optimizer. amsgrad (`bool`, defaults to `False`): Whether to use the [AMSGrad](https://hf.co/papers/1904.09237) variant of Adam that uses the maximum of past squared gradients instead. adam_w_mode (`bool`, defaults to `True`): Whether to use the AdamW variant. args (`object`, defaults to `None`): An object with additional arguments. min_8bit_size (`int`, defaults to 4096): The minimum number of elements of the parameter tensors for 8-bit optimization. percentile_clipping (`int`, defaults to 100): Adapts clipping threshold automatically by tracking the last 100 gradient norms and clipping the gradient at a certain percentile to improve stability. block_wise (`bool`, defaults to `True`): Whether to independently quantize each block of tensors to reduce outlier effects and improve stability. max_unorm (`float`, defaults to 1.0): The maximum gradient norm. rrr Nr rrrrrrrrrrrrrr rs rr zLAMB8bit.__init__Ls<^             r rTr r!rFTNr#r$Frr%r*s@rr,r,K3   < < rr,c8eZdZ dfd ZxZS) LAMB32bitc<t|d|||||d| | | | d y)a) 32-bit LAMB optimizer. Arguments: params (`torch.tensor`): The input parameters to optimize. lr (`float`, defaults to 1e-3): The learning rate. bias_correction (`bool`, defaults to `True`): Whether to apply bias correction to the first and second-order moments. betas (`tuple(float, float)`, defaults to (0.9, 0.999)): The beta values are the decay rates of the first and second-order moment of the optimizer. eps (`float`, defaults to 1e-8): The epsilon value prevents division by zero in the optimizer. weight_decay (`float`, defaults to 1e-2): The weight decay value for the optimizer. amsgrad (`bool`, defaults to `False`): Whether to use the [AMSGrad](https://hf.co/papers/1904.09237) variant of Adam that uses the maximum of past squared gradients instead. adam_w_mode (`bool`, defaults to `True`): Whether to use the AdamW variant. args (`object`, defaults to `None`): An object with additional arguments. min_8bit_size (`int`, defaults to 4096): The minimum number of elements of the parameter tensors for 8-bit optimization. percentile_clipping (`int`, defaults to 100): Adapts clipping threshold automatically by tracking the last 100 gradient norms and clipping the gradient at a certain percentile to improve stability. block_wise (`bool`, defaults to `True`): Whether to independently quantize each block of tensors to reduce outlier effects and improve stability. max_unorm (`float`, defaults to 1.0): The maximum gradient norm. rr"rr Nr r/s rr zLAMB32bit.__init__s<^              rr0r%r*s@rr3r3r1rr3N)bitsandbytes.optim.optimizerrrr,r3rrr7s2 9@ ?@ F= = @= = r