+ ܳjT^RIHt!RR]4t!RR]4t!RR]4tR#) )Optimizer2Statec6aa]tRt^toRV3RlltRtVtV;t#)LAMBc  :<\S V`RVVVVVV V V V R7 R#)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. 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_sizer __class__s&&&&&&&&&&&&&X/home/luca/screen-checking/.venv/lib/python3.14/site-packages/bitsandbytes/optim/lamb.pyr LAMB.__init__ s6X            ) MbP?Tg?g+?:0yE>rFT N?__name__ __module__ __qualname____firstlineno__r __static_attributes____classdictcell__ __classcell__r __classdict__s@@rrrs7 7 rrc6aa]tRt^CtoRV3RlltRtVtV;t#)LAMB8bitc  <V'd \R4hV R8wd \R4h\S V` RVVVVV^V V V R7 R#)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. Note: This parameter is not supported in LAMB8bit and must be False. 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. max_unorm (`float`, defaults to 1.0): The maximum update norm for trust-ratio clipping. Note: This parameter is not supported in LAMB8bit and must be left at the default 1.0. The 8-bit blockwise update does not implement update-norm clipping; it is honored by the 32-bit LAMB / LAMB32bit optimizers. z&LAMB8bit does not support amsgrad=Truer#zFLAMB8bit only supports max_unorm=1.0 (default value for compatibility)rrN) ValueErrorr r r rrrrrrrrrrr rs &&&&&&&&&&&&rr LAMB8bit.__init__Ds[\ EF F  ef f           rr rTrr rFTNr"r#r$r,s@@rr/r/CsB B rr/c6aa]tRt^toRV3RlltRtVtV;t#) LAMB32bitc  :<\S V`RVVVVV^ V V V R7 R#)a 32-bit LAMB optimizer. Arguments: params (`torch.tensor`): The input parameters to optimize. lr (`float`, defaults to 1e-3): The learning rate. bias_correction (`bool`, defaults to `True`): Whether to apply bias correction to the first and second-order moments. betas (`tuple(float, float)`, defaults to (0.9, 0.999)): The beta values are the decay rates of the first and second-order moment of the optimizer. eps (`float`, defaults to 1e-8): The epsilon value prevents division by zero in the optimizer. weight_decay (`float`, defaults to 1e-2): The weight decay value for the optimizer. amsgrad (`bool`, defaults to `False`): Whether to use the [AMSGrad](https://hf.co/papers/1904.09237) variant of Adam that uses the maximum of past squared gradients instead. adam_w_mode (`bool`, defaults to `True`): Whether to use the AdamW variant. args (`object`, defaults to `None`): An object with additional arguments. min_8bit_size (`int`, defaults to 4096): The minimum number of elements of the parameter tensors for 8-bit optimization. max_unorm (`float`, defaults to 1.0): The maximum gradient norm. rrNr r2s &&&&&&&&&&&&rr LAMB32bit.__init__s6R            rrr4r$r,s@@rr6r6s4 4 rr6N)bitsandbytes.optim.optimizerrrr/r6rrrr:s2 98 ?8 vC C L5 5 r