README Dataset abstract: The data to generate figures 3 – 8: 1. Excel file (Figure 3(a)): contains the attenuations of the feedback fiber for figure 3(a) ; 2. Excel file (Figure 3(b)): contains the P(9) absorption line for figure 3 (b); 3. Excel file (Figure 4(a)): contains optical spectra for different pump transitions in figure 4(a); 4. Excel file (Figure 5(a)): contains the CW pump power/output power and stability of laser output as a function of time for figure 5(a); 5. Excel file (Figure 5(b)): contains the stability of laser output as a function of time for figure 5(b); 6. Excel file (Figure 6): contains the Measured output power as a function of pump repetition rate for figure 6; 7. Excel file (Figure 7): contains pump power/output power at selected repetition rates shown in figure 7. 8. Zip file (Figure 8): contains radio frequency spectra (a), optical spectra (b) and time dependence (c) for the pump (blue) and the laser (red) at selected repetition rates shown in figure 8. Methodology: The fiber attenuation was measured by using the cut-back method. The Tungsten lamp was used as the whitelight source. A 300 mm focal length scanning spectrometer (Bentham Instruments TMc300) was used for all spectral measurements at mid-infrared wavelengths, which has a 300 lines/mm grating and a liquid-nitrogen-cooled InAs detector (Electro-Optical Systems S-010-LN4). The acetylene (12C2H2) P(9) absorption near 1530 nm was measured by recording the transmission of CW laser after 10 m gain fiber filled with 0.3 mbar acetylene gas. A thermal power meter (Ophir 3A-SH) was used to measure the laser power. The tunable laser diode (ID Photonics GMBH, CoBrite DX1, linewidth <100 kHz, maximum output power ~40 mW) was used as a source after amplification. All the pump powers shown in Figs 4 (a) and 6 were measured after the dichroic mirror with a thermal power meter (Ophir 3A-SH). Stability of laser output as a function of time in the fig 5 (b), was measured at 10 Hz sampling rate over more than one hour. A high gain detector (Thorlab PDA20H-EC) was used to directly measure the lasing power. The mode profile (inset Fig 5 (b)), was measured by using a two-dimensional scan with a short piece of feedback fiber (less than 3 m) across the output beam. The fiber was mounted on a 2D scanning translation device and connected with a high gain detector (Thorlab PDA20H-EC). The whole scan was controlled by LABView software. Time-dependent (Figs 8 (a & c)) measurements used a high speed HgCdTe detector (Vigo PVI-3.4-1×1- TO39-NO WINDOW-35, 1ns rise time) for 3.1 µm and a high speed InGaAs detector (Thorlabs DET01CFC) for 1.5 µm.A 350 MHz digital oscilloscope (Agilent InfiniiVision DSO-X-3032A) and 6 GHz spectrum analyzer (Agilent CSA Spectrum Analyzer) were used for measurements